Accessibility hierarchy and acquisition of English relative clauses by Urdu EFL learners

Abstract This study aims to examine the errors committed by Urdu EFL learners in acquiring English Relative Clauses (henceforth RCs) by putting to test the predictions of hypothesis proposed by Keenan and Comrie’s (1977) Noun Phrase Accessibility Hierarchy (NPAH). It aims to inspect whether the NPAH constraint for the acquisition of English relative clauses be applicable to Urdu L2 learners or not. In order to accomplish the purpose of the study total 80 participants were selected from the University of Azad Jammu & Kashmir. Data were collected through questionnaire that was consisted of two tasks: grammaticality judgment task and sentence combining task. The study uncovered the fact that Urdu learners commit a number of errors in the production of English RCs. These errors are pronoun retention, head noun retention, shift of noun function, non-adjacency to head noun and relative pronoun, preposition stranding, and deletion, selection of wrong relativizer, change of lexical item, change of syntactic pattern, and deletion of comparative particle. Urdu EFL learners’ hierarchy of acquiring English RCs is partially consistent with NPAH. The present study confirms the difficulty of direct objective relative clauses compared to subject relative clauses and indirect relatives in Pahari, which strongly supports Comrie (2007) hierarchy that shows that the performance of Urdu EFL Learners for first three positions of hierarchy is consistent with NPAH while their performance for the remaining three positions is not consistent with NPAH. Urdu learners find it easy to perform on OCOMP RCs which is placed at the end in NPAH. Urdu EFL learners find Oblique RCs more difficult compared to genitive RCs.


Introduction
The investigation in Second Language Acquisition (henceforward SLA) has been one of the central concerns of investigators over the previous 40 years (Goodluck & Tavakolian, 1982;Sheldon, 1974).Among the well-studied structures in SLA, the acquisition of RCs has been the focus of many linguists.The investigation of second language acquisition through different acquisition order approaches is of great prominence.The emergence of Universal Grammar acquisition led to the development of research on Typological universals.Initially, behaviorism psychological theory proposed language transfer as the primary factor influencing second language learning (Noor, 1994).Advocates of behaviorism, who emphasized the exclusive role of transferring knowledge from one's native language (L1), believed that learners would find it easier to acquire grammatical structures similar to their L1, while different structures would be more challenging and acquired later.On the other hand, Chomsky (1965) posited the existence of an innate faculty in the human brain that aids language acquisition.This innate faculty, known as universal grammar (UG) (White, 2003), consists of a set of grammatical principles present in all languages, which govern the construction of syntactic structures.Typological universals are considered to be the manifestation of these deep linguistic principles proposed by universal grammar.
One significant area of investigation in second language acquisition research revolves around the acquisition order of strategies used in forming relative clause structures.The Noun Phrase Accessibility Hierarchy (NPAH), introduced by Keenan and Comrie (1977), is a universal that has been extensively examined in this context.Additionally, other theoretical proposals have been put forward to determine and comprehend the order of difficulty/acquisition of relative clause structures across different languages.Two notable hypotheses used in studying the acquisition order of relative clauses in second languages are the Perceptual Difficulty Hypothesis (PDH) proposed by Kuno (1974) and the Dependency Locality Theory (DLT) proposed by Gibson (1998Gibson ( , 2000)).Both hypotheses focus on the complexity of cognitive processes involved in comprehending and producing relative clause structures.DLT identifies embedding and incompatible word order as sources of complexity and, consequently, difficulty in acquisition.It also suggests specific criteria for measuring the complexity of structures based on the distance between the basic noun phrase (NP) and the basic verb phrase (VP) in the main sentence.
It is important to note that these proposals emphasize the role of language-specific features in acquiring syntactic structures.However, their applicability to Urdu is limited due to its flexible word order, as it differs from the fixed word order systems on which these hypotheses are based.This study therefore, uses NPAH (1977) as a framework to study the acquisition order of relative clauses by Urdu EFL learners.
Researchers have investigated various aspects of the acquisition of English relative clauses and the accessibility hierarchy theory, including the order in which different types of relative clauses are acquired, the processing strategies employed by children, and the factors that influence the ease or difficulty of acquisition.Hawkins and Chan (1997) observe that while learners demonstrate overall competence in acquiring relative clauses, there are specific features within these structures that remain problematic or inaccessible.These features may include syntactic complexities or interpretive constraints related to the use of relative clauses.Tsimpli (1997)found that the acquisition of relative clauses in a second language follows a developmental sequence, with learners initially producing simpler structures and gradually progressing to more complex ones.Learners often struggle with the correct use of relative pronouns, word order, and the interpretation of ambiguous.Similarly (Felser et al., 2009) observed that learners may experience difficulty in processing object-extracted relative clauses due to the increased distance between the head noun and the relative pronoun.The literature suggests that the accessibility hierarchy may affect the learners' ability to process and comprehend different types of relative clauses.According to Ju (2014) Relative clauses are complex syntactic structures that play a crucial role in English sentence formation and comprehension.However, their acquisition poses challenges for non-native learners, particularly those from different linguistic backgrounds.He found that that L1 transfer plays a significant role in the acquisition of relative clauses by non-native EFL learners.
By extending the existing body of knowledge, this study aims to contribute to a more comprehensive understanding of how Urdu EFL learners acquire relative clauses within the framework of the accessibility hierarchy.Urdu RCs are different from English RCs in terms of syntactic structures and construction.Both the languages show the difference in relative clause formation.English relative clauses appear to the right of head noun, so English is a right-directed branching language.In contrast, Urdu primarily has both right and left direction branching, in which relative clauses pre modify as well post modify the head.Both of these differences can be a cause of the trouble in acquiring English RCs by Urdu L2 learners.The study seeks to investigate whether the Accessibility Hierarchy holds true for Urdu EFL learners and whether the linguistic differences between Urdu and English influence their acquisition process.

Problem statement
Despite the increasing importance of English as a global lingua franca, many Urdu EFL learners encounter challenges in acquiring complex grammatical structures, such as relative clauses.The Accessibility Hierarchy, which ranks the ease of processing different types of relative clauses based on syntactic and semantic factors, has been identified as a crucial factor in native language acquisition.However, the specific influence of the Accessibility Hierarchy on Urdu EFL learners' acquisition of English relative clauses remains underexplored.Therefore, this study aims to investigate the relationship between the Accessibility Hierarchy and the acquisition of English relative clauses by Urdu EFL learners, seeking to identify potential difficulties in learning these structures and provide insights to inform more effective language teaching approaches.

Research objectives
This study is an attempt: (i) To investigate and analyze whether Urdu speakers follow the patterns proposed by the Noun Phrase Accessibility Hierarchy when acquiring relative clauses in English.
(ii) To find out the type of difficulties in acquisition of RCs.

Research questions
The study addressed the following research questions: (i) Is the acquisition of English RCs by Urdu speakers consistent with the NPAH?
(ii) What particular difficulties do these learners experience in the acquisition of RCs?

Significance of the research
The significance of the study "Accessibility Hierarchy and acquisition of English Relative Clauses by Urdu speakers" lies in its potential to enhance the understanding of how Urdu speakers acquire relative clauses in English, which can have broader implications for second language acquisition and linguistic research.By investigating the accessibility hierarchy, which determines the ease of processing different types of relative clauses, the study sheds light on the cognitive mechanisms involved in language transfer and the challenges faced by Urdu speakers in acquiring this aspect of English grammar.These findings may aid in developing more effective language teaching strategies for Urdu speakers learning English and contribute to the field of linguistics by adding valuable insights into the relationship between language structures and second language acquisition processes.
By examining the performance of Urdu EFL learners in comprehending and producing English relative clauses, the study aims to shed light on the specific difficulties they encounter and provide insights into effective pedagogical approaches to enhance their relative clause acquisition skills.The findings of this study can contribute to the field of second language acquisition and language teaching by helping educators develop targeted instructional strategies to address the challenges faced by Urdu EFL learners in acquiring English relative clauses.Additionally, the study can provide a deeper understanding of the influence of language-specific factors on the acquisition of complex syntactic structures, offering valuable insights for researchers interested in cross-linguistic studies and linguistic typology.

Noun Phrase Accessibility Hierarchy (NPAH)
For investigating the order of acquisition in RCs structure, along with the studies that inspected the hypothesis of other theories, NPAH was the most common one.Keenan and Comrie (1977) suggested the Noun Phrase Accessibility Hierarchy (NPAH) after conducting their studies on 50 languages which, according to Fox (1987), is the strongest clarification of RC.Keenan and Comrie (1977) proposed a hypothesis Noun Phrase Accessibility Hierarchy (NPAH) that examined the different tactics, which are used in relative clause structures in terms of the variation in syntactic positions of the relativized noun.On the basis of markedness level, it also recommends the acquisition order of these structures.
After collecting the data from about 50 human languages Keenan and Comrie (1977) proposed the following hierarchical model for acquisition of relative clauses:

SUB>DO>IO>OBL>GEN>OCOMP
(1) SUB = Subject, as in (The girl who is beautiful is my cousin.) (2) DO = Direct object, as in (The guy that the puppy tad was intelligent.) (3) IO = Indirect object, as in (The teenager that I transcribed a document to was ill.) (4) OBL= Oblique, as in (The pen you are searching for is not your.) (5) GEN = Genitive, as in (The boy whose book was lost is smart.)(6) OCOMP = Object of comparative, as in (The lady whom I am taller than is talkative.) The accessibility hierarchy refers to the cognitive difficulty or complexity of processing different types of relative clauses.The hierarchy ranks relative clauses based on their difficulty level for learners, ranging from subject-extracted clauses as easier to object-extracted clauses as more challenging.The accessibility hierarchy hypothesis suggests that relative clauses closer to the beginning of the hierarchy are acquired earlier and with greater accuracy compared to those further down the hierarchy.These positions are not exactly same in all the languages, rather on the basis of markedness.These positions form a hierarchical model.The most extensively dispersed structure is considered as less marked structure.Less marked structure means it is more available which in turn is easier to be learned.More marked structures are not dispersed extensively that causes difficulty in learning.Consequently, Keenan and Comrie (1977) gave the name to this model as "Noun Phrase Accessibility Hierarchy" (NPAH)

Studies conducted under the Noun Phrase Accessibility Hierarchy (NPAH)
Several SLA studies have testified whether the RCs acquisition order by learners follows the NPAH or not.Among them Izumi (2003) conducted a study to investigate the acquisition order of RCS.The study inspected three hypotheses which were NPAH (Keenan & Comrie, 1977), PDH (Kuno, 1974).A total of 61 native speakers of 12 languages were taken, who were learning English as their second language.The study concentrated on SUB, DO, and OBL structures for NPAH.Results of this did not match with the NPAH Hierarchy, the researchers found that the learners exhibited similar processing patterns for both subject and object relative clauses.This outcome challenges the notion that subject relative clauses are inherently easier to comprehend than object relative clauses.Diessel and Tomasello (2005) is of the view that Relative clauses are complex syntactic structures.Due to structural complexity, these clauses are acquired relatively late by EFL learners.Through sentence combining task, he examined the accuracy construct.A total of 50 students were taken who were learning Japanese as their second language and their native language was Cantonese.Students were from intermediate level and from advanced level.The study included SU RCs, DO RCs, and OBL RCs.The results disclosed that SU and DO had the same level of difficulty while OBL was more challenging.Contrary to the NPAH, they found no significant difference in the acquisition of SU RCs and ORCs.Similarly, Ozeki et al. (2007) conducted two studies to predict the difficulty order of RCs by Japanese second language learners.In the first study, the construct of frequency was analyzed.A total of 90 learners were taken in the study and they belonged to different proficiency levels.Data was taken in the form of oral production by these 90 learners.Results indicated that the learners even from the low proficiency level were able to produce direct object RCs and oblique RCs.So these structures are not more difficult than SU RCs.Gibson (2010) inspected Chinese RC handing out different viewpoints.It concentrated merely on L1 acquisition not on acquisition of L2, so the focus of the study was native speakers.It concentrated on the understanding of RCs not on the construction of RCs.The study was conducted with 40 participants and their reading speed was determined by software.Surprisingly, their results did not support the traditional NPAH pattern.The participants showed comparable proficiency in comprehending and producing object RCs compared to subject RCs.SUB RCs.Outcomes indicated that contestants find it easier to read object RCs compared to subject RCs.SUB RCs were difficult to read hence this rejected the NPAH.Ju (2014) conducted research for predicting difficulty order by testing NPAH and markedness.The study was conducted on native and second language learners of Korean.The study included four groups.One group was the native speaker of Chinese language, second was the native speaker of Japanese, third group consisted of the native speaker of English, and all of these groups were learning Korean as their second language.Fourth group was the native speaker of Korean language.Findings revealed that there was no substantial difference among all groups performance for SU and DO RCs.For Korean first and second language acquisition, NPAH and markedness cannot be applied; hence NPAH and markedness are not universal.Ju's findings challenged the universality of the NPAH, as languages with dissimilar word orders showed contrasting relative clause preferences.This study highlights the potential influence of typological differences and language-specific factors on relative clause acquisition In Madsen's (2015) study, the Noun Phrase Accessibility Hierarchy (NPAH) was examined in relation to EFL learners' acquisition of relative clauses (RCs).Participants were given a clausecombining test and a gap-filling test.In the clause-combining test, they had to insert a second clause as a relative clause into pairs of independent clauses with a common referent.In the gapfilling test, they were required to insert the appropriate relative pronoun into matrix clauses.The results indicated that despite the similarities between Danish and English in genitive relativization rules, Danish learners faced difficulties with this type of relativization.However, the study found that the difficulty order of RC types did not align with Keenan and Comrie's (1977) NPAH.
Hayat (2016) conducted a study that focused on the challenges faced by Arab adult EFL learners in acquiring English restrictive relative clauses.The study also explored factors influencing the acquisition process.To identify potential acquisition problems, an acceptability judgment test was administered to 100 Arab EFL learners.The results indicated that the participants accepted the use of resumptive pronouns and showed a preference for overt relative pronouns over covert ones.He argued that the factors, such as learners' native language structures and exposure to relative clauses in the target language, may play a more significant role in determining the ease of acquisition.
In his study, Maani (2019) investigated how Jordanian EFL learners acquire English relative clauses.Additionally, the study aimed to determine the applicability of the Noun Phrase Accessibility Hierarchy (NPAH) proposed by Keenan and Comrie (1977) to Jordanian EFL learners in relation to the acquisition of relative clauses.Overall, the findings indicated that Jordanian EFL learners demonstrate proficiency in producing relative clauses.However, their performance was influenced by their proficiency levels, with advanced learners outperforming intermediate learners.Moreover, the results revealed that the NPAH effect is not applicable to Jordanian EFL learners, regardless of their proficiency levels.Koçak (2020) examined the level of recognition of relative clauses among Turkish tertiary level EFL learners.The study involved 30 philology students studying at the Preparatory School of Hacettepe University, who were given both a pre-test and a post-test.The findings revealed that while most participants were proficient in identifying relative clauses, their recognition level varied significantly across different types of relative clause constructions.Additionally, the results demonstrated that explicit instruction greatly facilitated the acquisition of relative clauses.
Albikri and Jarrah (2022) investigated the acquisition of relative clauses in Arabic and English by second language (L2) learners who are native speakers of English and Arabic.The study aimed to assess the extent to which the Noun Phrase Accessibility Hierarchy (NPAH) (Keenan & Comrie, 1977) and the Markedness Differential Hypothesis (MDH) (Eckman, 1977) account for the acquisition of relative pronouns in English and Arabic among L2 learners.A total of twenty Arabicspeaking learners of English and twenty English-speaking learners of Arabic were selected as participants.The findings revealed that Arabic learners of English were able to comprehend and produce relative pronouns successfully, although their performance was influenced by the Arabic system of relative pronouns.Notably, they exhibited better proficiency in producing "who" compared to "whom".Moreover, the study indicated that the performance of Arabic learners of English varied depending on the task type, with the Sentence Combination Task being more challenging than the Multiple-Choice Task.Conversely, English-speaking learners of Arabic (i.e., the L2 Arabic group) demonstrated ease in producing and comprehending Arabic relative pronouns, despite interference from their native language.Overall, the results supported the predictive abilities of the NPAH and the MDH in identifying the errors made by L2 learners of Arabic and English.
Different studies find that in both first and second language (L1, L2), the learning difficulties of different types of RCs are constant with NPAH (Gass, 1979;Albikri and Jarrah (2022) etc.).On the other hand, findings of several L2 acquisition studies, containing Korean (Ozeki et al., 2007), Cantonese and Japanese (Jeon & Kim, 2007;Maani, 2019)) are not consistent with the NPAH.These studies highlight the need for a more nuanced and comprehensive approach to studying language acquisition, one that takes into account individual differences and the complexity of linguistic input.Future research should continue to explore how the Accessibility Hierarchy operates across different languages and investigate potential pedagogical implications for language instruction and intervention programs.Understanding the intricacies of relative clause acquisition contributes to a broader comprehension of language development and human cognition.

Methodology
The study employed quantitative approach.The targeted population of the study consisted of 150 students of Department of English University of Azad Jammu & Kashmir Muzaffarabad.There were total 150 EFL learners in the department, out of them only 100 were willing to take part in the study.During the data collection process 20 participants either didn't attempt the tests or left incomplete.Therefore, the final sample for the study consisted of 80 participants, 30 males and 50 females as the ratio of the female students was higher than the male one in the department.All the subjects were BS English students and native speakers of Urdu who have been studying English as second language from last 5 years.Typically, the concept of relative clauses is introduced during the Intermediate levels of language learning.This ensured that the participants possessed prior knowledge of the structure of relativization that the tasks aimed to assess.

Data collection procedure
In addressing the objective, the present researcher employed two tests, sentence combination test and grammaticality judgment test, which were borrowed from Izumi (2003).The two tasks were administrated on two separate days at the Department of English, University of Azad Jammu and Kashmir, towards the end of the academic year 2021-2022.Recognizing that the test environment can influence performance (Backman, 1990), the testing was conducted in a familiar location for the participants and utilized a method they were accustomed to, namely pen and paper.The research participants were gathered in a single classroom, where the researcher served as their teacher, and administered all the tasks during the regular class time for the participants.

Grammaticality judgment task
The Students were asked to give responses concerning the grammaticality of 30 English relative clauses, 15 of which were ungrammatical and 15 were grammatical.The subjects were also asked to correct those sentences that are marked as ungrammatical by them.This task was intended to capture the level of acceptance of Urdu speakers towards several kinds of English relative clauses, in which different grammatical functions of NP are relativized.Therefore, six main groups of sentences were used in the grammaticality judgment task containing the nouns at all different position identified by Keenan and Comrie (1977).

Sentence-combining task
In order to check whether or not the order of hierarchy goes consistent with NPAH in case of Urdu EFL learners, only grammaticality judgment task was not enough because in this task too much productivity was not involved.So to check the productivity of English relative clauses another task that was sentence combining task was carried out.This task was intended to check the hierarchy order as well as to find out the difficulties faced by learners.
The Students were given 14 pairs of sentences and were asked to combine them to form English sentences with a set of relative pronouns.The main aim of this task was to check whether learners find it easy to produce English relative clauses or not.We also examine the type of errors they commit during the production of English RCs.

Data analysis
To achieve the objective, the current study utilized two tests: the sentence combination test and the grammaticality judgment test, which were adapted from Izumi's (2003) research.

Grammaticality judgment task on English relative clauses
The data obtained from this task was utilized to determine the frequency (represented as a percentage) of acceptance for various types of English relative clauses based on the NP functions of the NPAH.The percentage calculation was used to assess participants' ability to judge the grammaticality of English relative clauses accurately and provide appropriate corrections or alternatives for ungrammatical sentences.Particular attention was given to cases where respondents marked a sentence as ungrammatical, to assess whether they provided the correct alterations or not.

Sentence-combining task on English relative clauses
The researcher manually examined the collected data from this task, specifically focusing on errors related to the formation of relative clauses.Only those errors were tallied and documented.This task was primarily qualitative in nature, aiming to examine learners' production of different types of relative clauses.The researcher identified and categorized all errors pertaining to the production of English relative clauses using thematic coding, and percentages were employed for further analysis.

Results
This section presents the results of Grammaticality judgment task and sentence combining task.These tasks were designed to answer the research questions.

Results of grammaticality judgment task
The main aim of the task was to check the receptive or intuitive knowledge of the learners towards English relative clauses, that either they can clearly identify the relative clauses are grammatical or ungrammatical.Results from grammaticality task are shown in Table 1.
As the data in table shows, there is high frequency of correct identification on subject relative clauses.64.25% respondents correctly identified whether subject relative clauses are correct.After subject relative clauses 62.5% respondents made correct identification on direct object relative clauses.Then, 59.5% respondents gave correct responses on indirect object relative clauses, and after, those 58.4% respondents clearly identified object of comparative relative clauses.56.75% learners make correct identification on genitive relative clauses and 33.75% learners identified on Oblique relative clauses that whether these are grammatical or ungrammatical.95.75% respondents correctly identified ungrammatical subject relative clauses, 93.5% respondent gave correct responses for direct object relative clauses, 92.25% respondents made correct identification for indirect object relative clauses, 82.5% respondents correctly identified oblique relative clauses, 85.5% accepted the correct genitive relative clauses, and 86% learners gave correct response for object of comparative relative clauses.
As the data in above table shows, the frequency of learners in accepting ungrammatical sentences as grammatical is very low on subject RCs, that is 5.25%.Then, 6.5% learners accepted the ungrammatical sentences as grammatical on direct object RCs.7.75% learners accepted the ungrammatical indirect object RCs.14% learners accepted the ungrammatical sentences on object of comparative RCs.14.5% learners accepted ungrammaticality on genitive relative clauses.17.75% learners accepted the ungrammaticality on oblique relative clauses.
This was a judgment task to check the knowledge of learners so that they can identify that whether the sentences are grammatical or ungrammatical.The next step was to make the correction on those sentences that were identified as ungrammatical.Results are shown in Table 2.
We see that many learners clearly identified grammatical and ungrammatical clauses.The data shows that only 26.51% respondents gave the correct alterations to the ungrammatical sentences to make them grammatical in subject relative clauses.Then, 24.5% respondents found it easy to  The data in the table above show that 15.5% respondent made inaccurate corrections on subject relative clauses after identification that sentence is incorrect.After subject relative clauses.17% respondents gave incorrect alterations to direct object relative clauses.18.95% learners found it difficult to give correct alteration on indirect object relative clauses.22% learners found difficulty on object of comparative relative clauses for correcting these clauses.25% respondents found it difficult to make correct corrections on genitive relative clauses.27.5% learners find it tough to give accurate correction on oblique relative clauses and this percentage is high compared to other relative clauses.
Many learners, after identifying that sentences are incorrect, they did not give any correction on those sentences.The data in the table show that there is a low frequency of respondents who did not give any correction on subject relative clauses after identifying that sentence is ungrammatical.22.25% respondents gave rejection without giving any correction.24%learners find it difficult after identifying to give it correction on direct object relative clauses.25% learners identify without any correction on indirect object relative clauses.The frequency of respondents is high on object of comparative relative clauses compared to direct, indirect and subject relative clauses, that are 28% who did not give any correction on object of comparative RCs.32% did not give corrections on genitive relative clauses.The percentage of learners is high on oblique relative clauses who find it difficult to give accurate corrections.35% respondents fail to give any correction on oblique relative clauses.So many learners, after identifying that sentence is not correct, did not give any corrections.Some learners also accept the ungrammatical sentences as grammatical.They did not identify that sentences are ungrammatical.

Results of sentence combining task
Sentence combining task was conducted to check the students' performance on productive part.It was intended to check the student abilities to form relative clauses with the appropriate use of relative pronouns by combining two sentences.
The data in Table 3 show

Retention of pronoun
EFL Learners retained the resumptive pronoun while producing relative clauses in sentence combining task.As the data in Table 2 shows that, none of the respondents retained the anaphoric pronoun in subject and direct object relative clauses.Only 3% respondents retained the resumptive pronoun in object of comparative clauses while 9% respondents retained pronoun in genitive relative clauses.16%respondents retained pronoun in oblique relative clauses while 25% respondents retained pronoun in indirect object relative clauses.The learners found more difficulty in indirect object relative clauses compared to other relative clauses.
So, pronoun retention is cross linguistically unmarked, as more learners apply it while producing relative clauses.However, Urdu learners retained the anaphoric pronoun may be due to the native language influence.Urdu uses correlative marker "vo", which somehow resembles with English anaphoric pronoun and gives reference back to Head noun.

Retention of head noun
Another problem that learners faced while producing relative clauses was the retention of head noun in the relative clauses.Retention of head noun means that head noun was retained in the relativized noun phrase.The retention of head noun was found only in production data.Retention of same element within the sentence is odd and unnatural.Such errors are rare but not impossible.
The data show that very few learners retain the head noun in subject and direct object relative clauses.Then some learners retain head noun in object of comparative and genitive relative clauses.The ratio of retaining head noun is higher in oblique and indirect object relative clauses.
The retention of head noun is also due to the mother tongue influence because Urdu retains the head noun with the correlative in relative clauses.

Shift of noun function
Another difficulty that students face while producing the relative clauses was the shift of noun function.Students deliberately switch the order of two sentences during embedding, which serves as a strong evidence of avoidance.Most of learners shift the function of noun in order to produce the relative clauses that they find easy instead of producing the targeted relative clauses.None of the students shifted the function of noun in subject and direct object relative clauses, which suggests they found easy to produce subject and direct object relative clauses, which supports the NPAH.50% learners shifted the function of noun in indirect object relative clauses.This shows that learners simply avoid producing oblique relative clauses.Obviously deviated form is easy to produce as the noun phrase serves as direct object of matrix clause and the subject of relativized clause.95% learners failed to produce genitive relative clause due to its complexity.They shifted the noun function and made other clauses instead of genitive clause.The learners did not produce genitive relative clause, they found it difficult and shifted the noun function and made the subject relative clause.Thus, the major difficulty for learners was to produce the indirect object, oblique, genitive, and object of comparative clauses, and they shifted NP function and produced the subject and direct object relative clauses, that again goes in line with NPAH.

Non adjacency to head noun and relative pronoun
Another difficulty which learners find while producing relative clauses was that they produced the relative clauses which were not adjacent to head noun and relative clause.English relative clauses are directly adjacent to head noun, which means there is no other element between the head noun and the relative marker.Learners produced relative clause, which were not adjacent to head noun and relative pronoun.Only 3% of learners produced extra posed relative clauses in subject relative clauses.Instead of making the relative clause with adjacent head noun and relative pronoun learners produce the extra posed relative clauses.40% of learners produced the extra posed direct object relative clauses in which there was a gap between the head noun and the relative pronoun.In direct object relative clauses learners also generate the extra posed relative clauses.70% of learners produced the extra posed indirect object relative clauses.Most of the learners found difficulty in producing correct indirect object relative clauses.25% respondents produced extra posed oblique relative clauses.28% of learners produced genitive relative clauses in which head and relative pronoun were not adjacent.10% of respondents produced the object of comparative clauses in which the head noun and the relative pronoun were not adjacent and they produce extra posed relative clauses.The results indicate that most of learners had difficulties to produce correct relative clause, because they mostly produced direct and indirect object extra posed relative clauses.Very few learners produced subject extra posed clauses.In oblique, genitive and object of comparative relative clause, the issue of non-adjacency is not higher because most of the learners failed to produce these relative clauses, and they do shift the function of noun phrase as we have discuss earlier.The learners produced the extra posed relative clauses as it is easy to produce the relative clause with right embedding compared to produce relative clauses with center embedding.However, non-adjacent relative clauses did not hinder the semantic interpretation of sentences.Non adjacency of head noun and relative pronoun may be due to the mother tongue because Urdu relative clauses are sometimes not adjacent to head noun.

Selection of wrong relative marker
Another problem that the learners faced was the appropriate selection of relative pronoun to make the relative clauses.As we have seen, students did not find any difficulty in choosing right relative pronoun while making subject relative clauses.Only 2% of students uses wrong relative marker while making direct object relative marker.
20% of learners found it difficult to use right relative marker in making indirect object relative marker which is higher compared to other relative clauses.Only 7% of learners used the wrong relative pronoun in oblique relative clauses.10% of learners used incorrect relative pronoun in genitive relative clauses.Only 5% of students choose wrong relative marker in making the object of comparative clauses.

Change of lexical item
Another error that learners committed was the change of lexical item while constructing relative clauses.The change of lexical item was very rare in data.8% of respondents changed lexical item in genitive and 14% of learners changed the lexical category in indirect object relative clauses.

Change of syntactic pattern
Another error that learners committed was the change of syntactic pattern when producing relative clauses.However, learners only changed syntactic pattern in genitive and indirect object relative clauses, which is another indication of difficulty in genitive and indirect object relative clauses.Only 10% of learners changed the syntactic pattern in genitive relative clauses and 20% changed the syntactic pattern in indirect object relative clauses.Thus, the percentage of changing pattern is not very high.

Preposition stranding and deletion
Some learners deleted and stranded the preposition during the production of relative clauses.However, the percentage of preposition stranding and deletion is low.Only 15% of learners do that in indirect object and 20% committed this error in oblique relative clauses.They performed well for remaining positions.

Deletion of comparative particle
Another error that learners committed was the deletion of comparative particle.Only 10% of students deleted comparative particle in object of comparative relative clauses.The percentage of this error was not high.

Discussion
The analysis indicates that a large number of learners correctly identified grammatical and ungrammatical clauses.It means their intuitive knowledge about the target language is good.They perform well in receptive area but when it came to their productive part, the percentage with the accurate correction is very low compared to receptive knowledge.Many learners, after correctly identifying the grammatical and ungrammatical statement, did not give accurate corrections.The results from grammaticality task show it's easy for the learners to perform on subject relative clauses which support NPAH.After subject relative clauses, learners find it easy to deal with direct object relative clauses.They then find it easy to perform well on indirect object relative clause, which is again consistent with NPAH.After indirect object RCs, they perform well on object of comparative RCs.After indirect object RCs learners find it easy to perform well on object of comparative clauses and then genitive relative clauses.Learners find it more difficult in identifying whether the sentence is grammatical or ungrammatical and also in giving an accurate alternative to make them grammatical in oblique relative clauses.
There are several errors that Urdu EFL learners committed while constructing different English relative clauses in sentence combining tasks.These are retention of pronoun, head noun retention, shift of noun function, non-adjacency to head noun and relative pronoun, preposition stranding and deletion, selection of wrong relativizer, change of lexical item, change of syntactic pattern, and deletion of comparative particle.Results of sentence combining task show that there is divergence between the receptive and productive data.The percentage of respondents using presumptive pronoun in sentences is low compared to grammaticality judgment task.
After the results of both tasks that are grammaticality judgment task and sentence combining task, this work finds out that the Urdu EFL learners perform well in subject, direct object, indirect object relative clauses but they did not perform well in oblique, genitive, and object of comparative relative clauses in both the tasks.They find it easier to produce subject, direct object and indirect object relative clauses instead of producing genitive, oblique, and object of comparative relative clauses which are placed low in hierarchy according to NPAH.It is therefore obvious that when the ratio of production of these relative clauses was low, learners found it difficult to produce these clauses.This goes consistent with NPAH hierarchy.The learners find difficulty in indirect object relative clauses that are placed low in NPAH hierarchy.So the acquisition hierarchy of English RCs by Urdu EFL learners after the results of both tasks is: SUB>DO>IO>OCOMP>GEN>OBL So this hierarchy is partially consistent with the NPAH because Urdu EFL learners perform well on subject, then direct object and then indirect object relative clauses, but the last three positions are changed in case of Urdu EFL learners find less difficulties in object of comparative clauses compared to oblique and genitive relative clauses, and they find it more difficult to perform well in oblique relative clauses, which is not consistent with NPAH.The object of comparative relative clauses takes the last position in NPAH while it takes fourth position in the suggested hierarchy for Urdu EFL which is inconsistent with NPAH.Learners find oblique relative clauses most difficult, and they did not perform well in oblique relative clauses, so it takes last position in case of Urdu EFL learners, which is not consistent with NPAH, because it is at fourth position according to NPAH.These findings are partially consistent with the previous research of Pavesi (1986), who shows that there were two overturns: between indirect object relative clauses and oblique relative clauses, and between genitive relative clauses and object of comparative relative clauses.Pavesi (1986) sets up that OCOMP was learnt earlier than the less marked position GEN.Same is the case with Urdu EFL learners.GEN was not easier for respondents than the most marked position OCOMP according to NPAH.The findings of this study show that there is inverse order of OCOMP and OBL, while GEN is constant at the position.

Conclusion
This study concludes that hierarchical order of acquisition of English RCs by Urdu-speaking learners is partially consistent with the NPAH.The first three positions of English RCs are the same as those of NPAH.Urdu respondents found it easier to perform well on subject relative clauses, and then they performed well on direct object RCs.After direct object RCs, respondents perform better on indirect object RCs.The rest of the hierarchy is not consistent with NPAH.Urdu learners find it easy to perform on OCOMP RCs, which is placed on last according to NPAH.Then Urdu respondents find it easy to perform better on genitive RCs compared to oblique RCs.Urdu respondents find Oblique RCs more difficult, which is why placed on last according to the findings of this study.So the answer of the first research question on whether or not the Urdu EFL learners' acquisition hierarchy is consistent with NPAH, is that it is partially consistent with NPAH.It is not completely consistent with NPAH.The study also identifies different types of errors that respondents commit during the production of English RCs.These are retention of pronoun, head noun retention, shift of noun function, non-adjacency to head noun and relative pronoun, preposition stranding and deletion, selection of wrong relativizer, change of lexical item, change of syntactic pattern, and deletion of comparative particle.Each of these is a problematic area amongst our subjects in making relative clauses.Thus, these findings state the answer of second research question.The present study proposes the following hierarchy of Relative clause acquisition by Urdu EFL learners:

SUB>DO>IO>OCOMP>GEN>OBL
The current study provides a good insight into the order of acquisition of English RCs by Urdu EFL learners and points out the different difficulties that Urdu learners faced during the production of English RCs.The study suggests for a future research to see whether or not it would be effective for learners to teach them initially more marked structures rather than easier ones.

Funding
No funding was received to assist with the preparation of this manuscript.
6.The man whom I am taller than him is my roommate.Clearly good English (i) Miss Amina is kind to her students.
(ii) I always ask Miss Amina question on English grammar.
(i) The grammar book can now be purchased through the internet.
(ii) I bought the grammar book in the UK.
(i) The boy won the (ii) I am taller than the boy.
the following errors committed by the respondents in sentence combining task: Retention of Pronoun, Retention of Head Noun, Shift of Noun Function, Non Adjacency to Head Noun and Relative Pronoun, Selection of Wrong Relative Marker, Change of Lexical Item, Change of Syntactic Pattern, Preposition Stranding, and Deletion.
lion that carries the cow is horrible.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 8.The CD that you have been searching for was in the cabinet of a cupboard.whom the University has offered her a place to study is an outstanding student.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 10.The student whom the teacher has been arguing about the marks.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 11.The boy who his father has passed away has become very silent.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 12.The boy that Ali is smarter than won in the essay competition.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 13.The Pizza Hut which is in Islamabad is run by a movie star.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 14. Amina whom I invited her to join our dancing team comes from England.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 15.The little girl whom I teach German is very lovely.friend that I always talk to is frustrated by some interpersonal relationships.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 16.The man who his car I bought will leave Hong Kong very soon.girl whom Mary is prettier than her failed the test.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 18.The person that I live with her is my best friend.girl that you have waited for her an hour has a lot of admirers.uncle to whom I send cards every year always asks me to visit him in Canada.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 21.The topic that you should turn attention to it is related to housing problems.sister whose clothes I borrow has lot of admirers.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 23.The cousin who is younger than me graduated before me.chocolate which you like very much is very delicious.(a) Clearly good English (b) Probably good English (c) Probably bad English (d) Clearly bad English 26.The tutor that I am now sending the e-mail to him always dresses fashionably.