Comprehension of English Conditionals by Advanced Arabic-English Bilingual Speakers

The aim of this research is to determine the comprehension level of advanced Arabic L2 speakers of English conditionals, types I, II


Introduction
All natural languages are thought to have specific forms of conditional clauses that grammatically convey conditional meanings such as "reasoning about alternative situations, making inferences based on incomplete information, imagining possible correlations between situations, and understanding how the world would change if certain correlations were different" (Traugott et al., 1986, p. 5).English conditionals are made up of two clauses: the subordinate clause, the if clause, which includes the condition, and the main clause, the result clause (Fintel, 2009;Teschner & Evans, 2007).A conditional sentence's basic idea is that the action in the main clause cannot occur unless the condition in the subordinate clause is met.They primarily transmit the logical meaning 'if p then q,' but they can also be used to indicate other statements such as reasoning and imagining correlations (Traugott et al., 1986).
Although conditioning is universal (Comrie, 1986), English conditionals have long been regarded as one of the most difficult structures for second language learners to master (Celce-Murcia & Larsen-Freeman, 1999;Zhang, 2005).They are both grammatically and cognitively demanding.That is, they express many meanings through a variety of forms.According to Norris (2003) and Ramirez (2005), one source of difficulty derives from the fact that one circumstance is dependent on the fulfillment of another.Conditionals can also be used to communicate cognitive reasoning, counterfactual, possible and impossible, real and unreal, and hypothetical meanings.Others believe that the major problems in English conditionals are caused by forms, meanings, and time-tense correlations (Nekoueizadeh & Bahrani, 2013).
Native Arabic speakers learning English as a second language have been reported to have similar difficulties (Alkhawalda, 2013;Fareh, 2007;Rdaat & Gardner, 2017).These difficulties are mostly due to the relationship between tenses and the matching of verb forms in clauses.The majority of these pieces of research, however, are qualitative, giving an analysis of the participants' errors and their possible causes.Furthermore, no previous research has looked into the level of mastery of English conditionals among advanced speakers.Therefore, the aim of this paper is to explore the comprehension of English conditionals by advanced Arabic speakers of English as a second language (L2).This paper comprises the following sections: the next section is devoted to a review of the concepts and studies in the literature.Section three discusses the research methodology.Results and discussion are presented in section four.Section five presents concluding remarks along with the limitations of the research.

Literature Review
Conditionality has been a popular issue among researchers in recent decades.They have mostly focused on philosophy before gradually moving on to linguistics.The term 'conditional' has a variety of definitions in the literature.Concentrating on the presence of a conditional conjunction, Dancygier (2006, p. 1) defines conditional as "complex sentences, composed of the main clause (sometimes also called q, or the apodosis) and a subordinate clause (p, or the protasis)."The subordinate clause is introduced by a conjunction if in the English language.Likewise, a conditional sentence, according to Bennett (2003), is a sentence that has two clauses: the main clause and the subordinate clause, with the subordinate clause being introduced by a conditional conjunction.The example in (1) presents a typical English conditional sentence.
On the other hand, other research defines conditionals in terms of the causal relation between the two clauses without the presence of a conditional conjunction.Bhatt and Pancheva (2017) point out that "conditional structures are interpreted, in general terms, with the proposition expressed by the antecedent clause specifying the (modal) circumstances in which the proposition expressed by the main clause is true" (p.325).This definition includes other types of conditional sentences in which the presence of a causal relationship between the two sentences is required without the presence of a conditional conjunction.The examples in (2) illustrate how sentences (2a) and (2b) can be conditionally interpreted as "If you kiss my dog, you'll get fleas" and as "It would be nice if you do that" (Bhatt & Pancheva, 2017).

Conditional Sentences in English
To date, several studies have investigated conditional sentences in English.These studies have adopted different criteria which have resulted in different classifications.The first one is the degree of possibility.For example, depending on this criterion, Comrie (1986) classified conditionals into higher hypothetical and lower hypothetical, Huddleston and Pullum (2002) into Open and Remote and Palmer (1986) into Real and Unreal.According to Dancygier (2006) the terms open, non-counterfactual, and real conditionals refer to situations in which a speaker simply provides the related premises without expressing any opinions on them.In unreal, counterfactual and hypothetical scenarios, on the other hand, the speaker indicates a negative belief about the fulfillment of the proposition.Depending on this criterion, some traditional grammarians have classified English conditional constructions into three types, type I, type II, and type III (Azar & Hagen, 2009;Carter & McCarthy, 2006;Wishon & Burks, 1980) with each type being made up of different tenses.Other grammarians have classified English conditionals into four types, zero, future, present, and past conditionals (Cowan, 2008;Liu, 2011;Wu, 2012;Yule, 1998;Zhang, 2005).This paper follows the tripartite classification which is explained in the next section.

Conditional Sentence Type I
According to Azar and Hagen (2009) this conditional is used to describe the possibility of something happening in the future or in the present.A future action will only take place if a certain condition is met by that time.It is hard to anticipate whether the condition will be fulfilled, but the conditions look to be reasonable, so it will.The sentence structure in this type uses the simple present in the subordinate clause and the future tense in the main clause.For example, the sentence in (2) expresses the meaning that if he does what is required (studying) then he will be very likely to pass the test.Table 1 below shows the main structure of type I conditional.
(2) "If he studies hard, he will pass the test."

Conditional Sentence Type II
Type II conditional is used to imagine an event that is occurring now or in the future, to convey a situation, and to express the consequences of this hypothetical condition.Furthermore, this type is also used to describe unreal possibilities or impossible situations; it defines the possible actions that would happen if anything hypothetically occurred, contrary to fact in the present or future.In a nutshell, this conditional indicates a wish or hope that is against what is currently occurring or will occur in the future.The structure of the sentence in this type uses the past simple in the subordinate clause and one of the models would, should, could, and might are used in the main clause with the verb base form.The auxiliary is used in the subordinate clause with all persons (Azar & Hagen, 2009).An example of this type is given in (3) and Table 2 below summarizes the main sentence structure of this type.
(3) "If she were an artist, she would paint you."

Conditional Sentence Type III
Unlike types I and II, which describe real and unreal situations in the present and future, this conditional type describes events that occurred in the past.This implies that it is impossible to meet the condition, contrary to fact, since it has already occurred, or may have occurred, but will not occur again.It expresses a wish with the exception that it has no chance of coming true.In this form of sentence, the past perfect tense is used in the subordinate clause and the perfect conditional is used in the main clause (Azar & Hagen, 2009).An example of this type is given in (4) and Table 3 below summarizes the main sentence structure of this type.
(4) "If I hadn't lied to him before, he would have believed me."

Conditional Sentences in Arabic
There is a long history of research into Arabic conditional sentences, starting with traditional grammarians and progressing to modern linguistics.Several studies have investigated conditional sentences in classical Arabic (Alxos, 1993;Peled, 1992), Modern Standard Arabic (MSA) (Alotaibi, 2014;Ryding, 2005), and Arabic dialects (Alotaibi, 2014;Brustad, 2000).This section provides a brief introduction to MSA conditional sentences.
Arabic conditional sentences consist of an introducing particle in the protasis, each conveying a different degree of probability.The particles in, law, and ida are all roughly equivalent to 'if' in English.Holes (1995) discussed the differences between these three particles.The particle 'in' presents a proposition of possible realization but one that is uncertain.The particle 'ida' introduces a contradicting meaning that denotes that the time of the clause is uncertain but the proposition in its clause will inevitably occur; hence, it can be translated into English with either 'if' or 'when'.Finally, 'law' is considered as the contrary-to-fact particle (Holes, 1995, p. 238).
Most studies on MSA have divided conditional sentences into real and unreal.Ryding (2005) briefly discussed MSA conditional sentences and described law to be unreal conditional while in and id̠ a to be real conditionals.Similarly, Badawi et al. (2013) provided a descriptive analysis of modern Arabic grammar.They devoted a complete chapter to discussing conditional sentences in Modern Written Arabic (MWA).They proposed that MWA has three conditional particles, namely: the real conditional in 'if', the unreal conditional law 'if only', and id̠ a 'when.'They claimed that id̠ a, which was previously thought to have only a temporal interpretation, has become the most common conditional particle in MWA.In their classification, id̠ a has four interpretations, as a pure conditional, a temporal, a combination of temporal/conditional, and a hypothetical conditional as law.
Likewise, Alotaibi (2014) investigated conditional sentences in MSA and Taif dialect.Alotaibi's study aimed to show the similarities and differences in the use of conditional sentences in these Arabic dialects.He pointed out that the two varieties use the same conditional particles, in and law, and that they can be semantically divided into real: in and unreal: law.Nonetheless, he observed that in Taif dialect, the MSA standard marker sa-sawfa is replaced by the marker bi.This marker clearly indicates that a future reference 'will' is being made.He considered the particles in and law to be the main conditional particles, contradicting previous studies by ignoring the particle ida.
Other linguists disagree with the previous binary classification and proposed a different typology.Sartori (2019) conducted a syntactic-semantic analysis of conditional sentences in MWA.He analyzed a corpus of nine Modern Arabic novels published between 1963 and 2005.Sartori focused primarily on the particles ida and law as he noticed that the occurrence of the particle in in the data was 5.60% and almost disappeared.The analysis he proposed is based on an approach called a 'relational approach.'This approach assumes that the apodosis verb form and the conditional particle are related, and that this relationship aids in expressing time reference and modality meaning.He observed that the forms of the apodosis in the data were mainly in the perfect tense form (97.18%), which explains the importance of the apodosis.He claimed that the interaction between the two elements denotes the precise meaning of the sentence, which includes time reference and Modality meanings.Based on this procedure, Sartori proposed five semantic domains: present eventual, past eventual, potential, present unreal, and past unreal.
In a similar vein, Alfraidi (2017) described the binary classification-real vs. unreal-as "too broad" in his study.
He conducted an empirical analysis of the semantic and contextual structures of the conditional particles in, law and ida in MWA.The data he analyzed were collected from different sources, novels, plays, newspapers, politics, history and philosophy.In the analysis, he used the parameters proposed by Dancygier (2006): modality and time references, the connection between the two clauses, the discourse functions, and the interaction between the conditional particles and other linguistic elements.He classified the Factual/Non-Factual dichotomy into five semantic categories: factual, likely, open, tentative, and counterfactual.Even though it is not as common as other particles, the particle in is mostly used to express open conditionals; ida is dominant in the factual, likely, and open classes, while law is dominant in the tentative and counterfactual classes.

Previous Studies about the Acquisition of English Conditionals
There is a relatively small body of literature that is concerned with the acquisition of English conditionals by Arabic native speakers.These studies have demonstrated the low performance of Arabic L2 speakers of English when it comes to English conditionals.Fareh (2007) administered a test to senior and junior Arab students to examine their ability to recognize and produce English conditionals.The test was divided into two sections: one to assess students' recognition of conditionals' form, meaning, and function, and the other to assess students' production of conditionals.The test included four types of conditionals, type I, type II, type III, and implied.Implied conditionals are conditional sentences that do not have an overt marker (the 'if' component).The results revealed that students' performance was relatively low, with roughly 44% achieving the correct responses.Furthermore, the results demonstrated that students performed better in the recognition section than in the production section.Fareh (2007) concluded that the low performance of university-level Arab learners of English as a foreign language can be ascribed to methodological issues in English teaching at schools and universities.
Another study by Alkhawalda (2013) investigated the interpretation of English conditionals by native Arabic speakers.He conducted a grammaticality judgment task, in which students were asked to decide whether or not a sentence was correct and if not, they were asked to specify the ungrammatical part.One hundred university students participated in the task.All were majoring in English language.He also performed an analysis of the participants' corrections of the sentences.Analysis of the participants' responses revealed a lack of awareness regarding the different variations of conditionals.According to Alkhawalda (2013), conditional types are various and restricting them to only three types when teaching grammar is problematic.
Similarly, Rdaat and Gardner (2017) used a grammaticality judgment task to examine the most common errors made by Arabic speakers when using English conditionals.The study involved 20 people with varying educational degrees (bachelors, masters, and PhD levels).He also examined the corrections to the sentences made by the participants.The results showed that over half of the participants were unable to correctly determine the grammaticality of conditionals.Furthermore, it was shown that the most difficult aspects for Arab students were the linkages between time tenses and the matching of the verb forms in the two sentences.This demonstrated that conditional constructions are extremely challenging, even for individuals who are considered to have a high level of language skills.
Together, these studies have shown the complexity of the acquisition of conditionals and the difficulties that a foreign language learner faces.The objective of the present study is to explore the comprehension of English conditionals by Arabic speakers.This was accomplished through an online test consisting of 60 multiple-choice questions.The methodology used is described in the following section.

Methodology
Although the acquisition of English conditionals by Arabic speakers have been investigated (Alkhawalda, 2013;Fareh, 2007;Rdaat & Gardner, 2017), studies on the level of mastery of English conditionals among advanced speakers have been scanty.Therefore, this study investigates how well advanced Arabic L2 speakers of English have acquired conditionals on the level of comprehension.In more specific terms, the study attempts to answer the following questions: 1) To what extent have advanced Arabic L2 speakers of English acquired English conditionals at the comprehension level?
2) With which conditional type do advanced Arabic L2 speakers have the most difficulty?
3) Will Arabic L2 speakers of English perform better at the syntactic or semantic levels?

Participants
A total of 35 participants were recruited, of whom 19 were females and six males.All participants were Arabic native speakers who learned English as an L2.Eligibility criterion required participants to have attained 6.5 or higher in the International English Language Testing System (IELTS) or any equivalent in other formal tests such as the Test of English as a Foreign Language (TOEFL) or the Standardized Test for English Proficiency (STEP).Two participants were eliminated due to the fact that they attained an IELTS score lower than 6.5.This makes a total of 33 valid responses.The participants ranged in age from 25 to 57 (M = 29, SD = 6.56).They were of different educational levels, bachelors (n = 8), masters (n = 23), and PhD (n = 2) levels.

Materials
The task consisted of 60 multiple-choice questions.The materials were separated into 20 items for each form of conditional, so 20 questions for type I, 20 questions for type II, and 20 questions for type III.The 60 questions were divided into two sections: the first section consisted of 30 questions aimed at assessing the participants' comprehension of the syntactic structure and were in the form of sentence completion.The second section contained 30 inference questions about the meaning of conditional sentences.Thirty questions were adapted from Azzali (2019), while the rest were sentences taken from Azar and Hagen (2009) and rewritten as questions.Minor changes were made to fit the research sample and context.Furthermore, the materials were revised by two PhD holders in linguistics from Kind Saud University.

Procedure
The test was conducted online.Before beginning the test, participants were presented with the study objectives and instructions and they were asked about their demographic information such as gender, age, IELTS score, educational level, and major.They also expressed their consent for their responses to be used in the study.The test began with the first section which consisted of 30 questions that asked the participants to complete the conditional sentences by choosing one of four options.After completing the first section, the participants moved to the second section which included 30 questions about the meaning of the conditional sentences.All the questions were presented randomly and there was no time limit, but the test lasted about 30 minutes in most cases.

Data Analysis
The scores from the test were organized and analyzed.Each correct response was worth one point and incorrect responses were worth zero points.Therefore, 60 points were available for a full-mark score.A descriptive statistical test was conducted to attain the cutoff point.This is the point that splits the scores into high and low.Three statistical tests were then performed to test the following: • The participants' comprehension level of conditional (high vs. low).
• The difficulty of the conditional types (type I vs. type II, type I vs. type III, and type II vs. type III).
• The difficulty of the sections (syntactic structure vs. inferences of the meaning).

Results
To assess the participants' comprehension level of English conditionals, type I, II, and III, an online test was conducted.The test consisted of a total of 60 multiple-choice questions, 30 of which aimed to assess participants' comprehension of the syntactic structure while the other 30 aimed to assess participants' comprehension of the meaning.A score of 1 was given to each correct response and 0 for each incorrect response.A perfect mean score was therefore 60.Table 4 shows the mean score of the participants' performance in the test.Table 6 shows the performance of the participants in each type of conditional (type I, type II, and type III).Table 8 shows the participants' performance in the two sections (syntactic structure vs. inferences of the meaning).An independent sample t-test was conducted to compare the mean test score between syntactic structure section and inferences of meaning section.On average, the mean test score of the inference of meaning section was higher than the mean test score of the syntactic structure section.This difference, 2.60, was statistically significant, t (64) = 3.74, p < .001.Table 9 shows the results of the independent sample t-test.

Discussion
The test in this study investigated advanced Arabic speakers' comprehension of English conditionals, specifically types I, II, and III.The results of this study are discussed in the light of the research questions presented above.
The first question of the study focused on the overall performance of participants in the acquisition of English conditionals.The average score of participants' performance was relatively low (M = 46.18).This result was not as expected considering that all the participants are of advanced language level.Furthermore, the t-test demonstrated that participants' performance was low.The results did not appear to be statistically different from 46, despite the fact that the number of individuals who received a score equal to or higher than 46 was greater than those who received a lower score, n= 18 and 15, respectively.This finding is consistent with that of Fareh (2007), who demonstrated that the performance of Arabic university students was extremely low.Similarly, Rdaat and Gardner (2017) also found that nearly half of their participants were unable to accurately determine the grammaticality of conditionals.
The results from the test showed that some types of English conditionals are more difficult than others to learn and use correctly.Contrary to expectations, the conditional type II appeared to be the most difficult for advanced Arabic English speakers (M = 14.51).This finding is surprising and inconsistent with previous studies which showed that conditional type III was the most difficult (Fareh, 2007;Rdaat & Gardner, 2017).A possible explanation for this might be the discrepancy between form and meaning in type II conditionals, in which the past form is used to refer

Mean of scors
Types to the present or future.In type I, the present form is used to refer to the present or future and in type III, the past perfect is used to refer to the past.
The test was divided into two sections: syntactic structure and inference of meaning.The results revealed that participants performed better in the inference of meaning section than in the syntactic section.This indicates that participants had more difficulty dealing with syntactic structure than with meaning inference.Fareh (2007) showed that participants had more difficulties in recognizing the meaning of the conditionals than in recognizing the form.This differs from the findings presented here.It is difficult to explain this result, although it could be related to the fact that information about the meaning is easier to understand than information about the form.

Conclusion
This study investigated the comprehension of English conditionals type I, II, and III by advanced Arabic L2 English speakers.The study has shown that, despite achieving high scores in English formal tests like IELTS, the performance of advanced Arabic English speakers did not reach a high level.Furthermore, it was found that the difference in mean scores between conditional types was statistically significant.Conditional type II, with the lowest mean score, was the most challenging conditional type.Furthermore, an independent sample t-test demonstrated that participants scored better in the section on syntactic structure than in the reference of meaning.This indicated that the first section was more difficult than the second.However, several limitations need to be noted regarding the present study.First, the study focused only on the comprehension level and did not investigate the production level.Second, the study only examined the three basic forms of English conditionals, despite the fact that there are other variations.Lastly, the study is limited by the relatively small number of participants.All of these are specific limitations that can be addressed in future research.

Figure 1 .
Figure 1.The results of the comparison between the mean scores of the conditional types

Table 1 .
The structure of the Type I

Table 2 .
The structure of the Type II conditional

Table 3 .
The structure of the Type III conditional Meaning of the subordinate clauseVerb form in the subordinate clause Verb form in the main clause Unreal in the past Past perfect Would have + past participle

Table 4 .
Overall performance of participants in the testAs shown in Table4, the mean score of the participants was 46.18.A one sample t-test was conducted to compare whether the participants' mean was statistically different from a hypothesized value.The 50 th percentile (46) was used as the test value.The test results revealed that participants' mean score was not statistically different from 46, t (32) = 0.233, p = 0.81.The results of the one sample t-test are shown in Table5below.

Table 6 .
Performance of the participants by conditional types one-way ANOVA was conducted to compare the effect of conditional type on the participants' scores.The test questions were divided into three types: type I (n = 20), type II (n = 20), and type III (n = 20).The one-way ANOVA revealed that there was a statistically significant difference in participants' scores between at least two types (F (2, 96) = [6.66],p=[.002]).An LSD test for multiple comparisons found that the mean score was significantly different between type I and type II (p = [.007],95%C.I. = [0.362,2.243]).Furthermore, the test demonstrated that the mean test score differed statistically between type II and type III (p = [.001],95%C.I. = [-2.576,-0.696]).There was no statistically significant difference in mean test score between type I and type III (P = 0.48).Table7provides the results of the LSD test, and the mean difference between conditional types is shown in Figure1. A

Table 7 .
Results of the LSD test

Table 8 .
Performance of the participants according to test sections

Table 9 .
Results of the independent sample t-test