Why is “John Ran to the House” the Same as “John Went to the House Running” in Arabic?

The study explores how Arabic has the same conflation pattern characteristics as English even though it belongs to Verb-framed Languages. A focused-group approach is used to evaluate the effect of the first language (L1) and the potential role of proficiency in the acquisition of the English directional preposition ‘to’ with manner-of-motion to goal construction. One group consists of Saudi speakers at two levels of development; an intermediate and advanced proficiency levels; whereas, the second group (control group) comprises of English native speakers. Acceptability Judgment Task associated with video animation clips is designed to elicit participants’ judgments in the depicted event. Results indicated that the intermediate Saudi speakers accept the directional preposition ‘to’ with and without boundary-crossing event, as is the case of their L1, which was opposite for the advanced and native English speakers for the without boundary-crossing event. The advanced Saudi speakers accept the constructions of encoding the manner with the motion and expressing the manner as the complement depicting an appropriate description of the event, reflecting L1 influence. All the group’s judgment varies based on the acceptance to conflate the manner with the motion overexpressing manner as a complement in an event without boundary-crossing.


Introduction
The motion concept is universally acknowledged across all languages around the world based on its pervasiveness in the daily lives (Blackledge & Creese, 2017). Pavlenko and Volynsky (2015) state that the motion expression differs among languages. In the Second Language Acquisition (SLA), the acquisition of argument structures has been widely recognized, particularly in the area of the acquisition of manner-of-motion to goal constructions (Rothman & Slabakova, 2018). Talmy (1985) has intrigued the interest of researchers who investigated the realization of English manner-of-motion to goal constructions by different first language learners (L1s) (Cadierno & Ruis, 2006;Brown & Gullberg, 2010;Chen, 2005;Choi & Lantolf, 2008;Donoso & Bylund, 2015;Römer, O'Donnell, & Ellis, 2014;Vergaro & Iacobini, 2014). Talmy's lexicalization (word making process for expressing a concept) of the semantic primitives (concepts that are innately understood and cannot be expressed in simpler terms) in the motion event yields to cross-linguistic variation in the conflation (i.e., merging of information, texts, opinions, ideas, and more sets) of these primitives in the main verb. His work classifies the languages into Verb-framed languages, such as Spanish, Turkish and Semitic languages, which include Arabic and Hebrew and Satellite-framed languages, i.e., English and German (Note 1) topic.
Considering the manner verbs, Albaqami (2016) cited Slobin's (1996) study and highlighted two-tiered lexicon manner-of-motion verbs; where one exists in a routine task, i.e., jump, walk, fly, run and more. Whereas, the second one is more specific and expressive such as walking can be expressed specifically with words like wander, stroll and which for running can be jog. Concerning the second language learners, Lardiere (2009) has articulated that the L2 complete acquisition is based on their reassemble of the L1 features into their L2.
Based on Talmy's typology, the core characteristic of Verb-farmed languages is to encode the path with the motion, such as the following Spanish example in (1) from Slobin (1997).  Aske's (1989)  The question is what licenses the manner to be encoded with the motion in Arabic? Example (7) shows that manner conflate with motion with the directional prepositions ʔɨla 'to' in Arabic. However, Kabli (to appear) argues that the directional preposition ʔɨla 'to' is ambiguous and behaves differently from English 'to'. It implies the same meaning of English 'to' when it is associated with no boundary-crossing interpretation. Thus, it allows the conflation of the manner with the motion (7a and c). It denotes the interpretation of 'into' or 'onto' when it involves boundary-crossing (7b, c, and d). Due to the absence of these prepositions, Arabic adopts different means to express boundary-crossing. English and Arabic primary difference is the inability of Arabic to encode the manner with the motion when it comes to boundary-crossing event. However, Arabic allows expressing manners as a complement (example 8). Such construction is not ungrammatical in English (Note 2). Arabic allows another type of conflation, along with the lexicalization pattern. This is the conflation of the motion with the path in the verb root leaving manner to be expressed optionally as a complement (example 9).
(9) daxla alwaledu almanzil (jarian) Enter-I-PAST the boy the house (running) 'The boy entered the house (running)' The expression of manner separately from motion in Arabic is due to the notion of the boundary-crossing. Arabic allows expressing manner as a complement when the event involves displacement from one position to another. Özçalişkan (2013) investigates if the boundary-crossing constraint can be held responsible for the cross-linguistic differences. First, she examines if the Turkish speakers express path verbs in the free description of the boundary-crossing event, as is the case in their L1. Then, she imposed manner verbs when describing the event. She concludes that the boundary-crossing acts as a reliable test to determine the typology of any given language. Kabli (2013) investigates the motion event with the element of boundary-crossing, such as John went into the room running. She finds that their L1 still constrained Saudi speakers at the advanced and the intermediate stages of development. Both Saudi groups accepted the construction as an appropriate construction depicting boundary-crossing in English. This highlights that the argument structure is persistent in the acquisition due to the absence of the directional prepositions that involve boundary-crossing meanings in the L1. Iacobini and Vergaro (2014) examine a corpus study of manner-of-motion verbs in Italian. It shows that manner verbs are used equally in describing the motion event regardless of the presence or absence of the boundary-crossing specification.
As is the case in Arabic, Son (2009) examines Hebrew which belongs to Verb-framed languages and falls under the Semitic languages. She observes that Hebrew has two prepositions el and le that is equivalent to the English 'to.' They are used as directional prepositions and cannot co-occur with the stative verbs. Examples included are from Son (2007, p. 153) (Example 10).
(10) *yoni haya {el ha-/la-} xeder Yoni was All the/DAT.DEF-room 'John was to the room.' The Hebrew speakers use manner-of-motion verbs for directional goal interpretation (Son, 2009, p. 139 Indonesian is very similar typologically to Arabic and Hebrew. Son (2007) shows that the manner is encoded with motion and co-occur with the directional preposition ke 'to' for directional goal interpretation (example, 12). It is plausible to assume that Arabic has two ways to express the boundary-crossing event. One is to use the directional preposition 'to' that is associated with the boundary-crossing interpretation and expressing manner as a complement. The other way is to conflate the manner with motion with directional preposition ʔɨla, which leads to an ambiguous interpretation. The third type is to conflate the motion with the verb root path, which lies outside the study scope.
The present work explores the issue of the first language and the role of proficiency in the acquisition of manner-of-motion verbs with the directional preposition 'to' by L1 Saudi speakers at two phases of development. It is mostly concerned as to why Arabic encodes the manner with the motion, besides, to express the manner as a complement. It is assumed that since the Arabic 'to' allow two interpretations, the low level of proficiency Saudi speakers, unlike English, typically will transfer their L1 property in their acquisition of L2 construction. They will accept English 'to' as denoting both boundary-crossing and no boundary-crossing interpretations. Likewise, the L1 effect will appear in expressing manner as a complement. Both groups of Saudi speakers will allow manner separately from motion in the event of denoting boundary-crossing. However, it seems that proficiency will play a negative role in the acquisition of the relevant construction.
Since Arabic allows the two types of conflations, both groups of Saudi speakers will likely to conflate manner with motion and express manner as a complement for boundary-crossing as a proper description of the event in English. In contrast, Saudi speakers will conflate manner with motion, but not express manner separately from motion in the event without boundary-crossing since this structure, as typically realized in Saudi Arabic.
In light of the above predictions, the present study aims to address the following research questions: Q1: If L1 plays a role in the acquisition of L2 argument structures, will the intermediate Saudi group accept 'to' with an event depicting boundary-crossing the same as without boundary-crossing like their L1 and unlike English as are in the following cases: a. The man climbed to the tree house (i.e., into the tree house) b. The man climbed to the tree house (i.e., to the tree house) a. The bird flew to the cage Q4. If proficiency plays a role in the acquisition of manner-of-motion to goal construction, will advanced Saudi group disallow 'to' with boundary-crossing event and disallow to express the manner as a complement in the event with boundary-crossing, as follows?

Q2: Will both groups of Saudi speakers accept expressing manner as a complement with 'to' in the event
a. The man swam to the cage (+boundary-crossing event) b. The man went to the cave swimming (+boundary-crossing event)

Study Design
A mixed study design is adopted as it presents the data in the statistical form while providing a comprehensive analysis of the findings. The focus group approach was used for assessing the lexical difference between the two languages. Moreover, this approach is also instrumental in overcoming that prevail in a specific approach which facilitates drawing a holistic picture of the research problem.

Study Participants
An anonymous approach was adopted for performing the study. In it, the participants were recruited by word of mouth in a Saudi Organization. Three groups were formed, where the control group composed of 15 English native speakers while the other two groups were of L1 Saudi participants and L2 English participants at two proficiency levels. There were 16 Saudi speakers at an advanced level of proficiency and 17 Saudi speakers at an intermediate level. Saudi participants were classified as an intermediate level based on paper and pencil Oxford Placement Test. However, the advanced proficiency of Saudi participants who were employed at the institute were enrolled in MA and Ph.D. programs in the USA and the UK. They reported their IELTS and TOFEL test scores as acceptable for admission in the universities. However, the mean score of IELTS was 7.5 whereas the mean score of TOEFL (IBT) was 100.5.

Questionnaire
An Acceptability Judgment Task accompanied with video animation clips examined participants' responses to the task items. Participants watched the animation pictures on the screen of the laptop and marked their responses on the answer sheets. The responses were based on 3-point Likert scales (i.e., 1 acceptable, 2 not sure and 3 unacceptable). They were requested to select acceptable if the statement describes the motion they saw on the screen. However, if they found that the animation picture mismatched the statement, unacceptable was to be selected. If participants were unsure if it was a proper English statement, they select not sure. Manner verbs used in the task were selected based on Levin's (1993) English verbs classification. The task was administered individually to all participants by appointment or during their breaks.
Participants were introduced to two examples of the animation clips at the beginning of the task. The setting of the animation clips involves describing two situations. The first situation designed as an individual crossed the boundary to inside another place. There were three choices described in this situation. The first choice involved the directional preposition 'to' with the manner incorporated with the motion in the verb. The second statement expressed the manner separately from the motion with directional preposition 'to' whereas the last choice was a distractor (example, 13). In contrast, the second situation implies an event without boundary-crossing, i.e., there was no displacement from one place to another. Again, there were three choices describe the motion event. The first statement includes the directional preposition 'to' with the manner conflated with motion. The second ijel.ccsenet.org International Journal of English Linguistics Vol. 10, No. 3; situation involves the directional preposition 'to' with expressing manner as a complement while the last choice was a distractor (example, 14). Therefore, there were six animation pictures with boundary-crossing motion events while the other six pictures were designed without a boundary-crossing motion event. The choices were 24 statements related to the study while the remaining 12 choices were distractors. Also, there were ten animation clips functioned as fillers.
(13) A. An event involving no B-C animation clip The boy climbed to the tree house Acceptable Not Sure Unacceptable The boy went to the tree house climbing Acceptable Not Sure Unacceptable The boy climbed to the top of the tree house Acceptable Not Sure Unacceptable

(14) An event involving B-C animation clip
The boy climbed to the tree house Acceptable Not Sure Unacceptable The boy went to the tree house climbing Acceptable Not Sure Unacceptable The boy climbed to the top of the tree house Acceptable Not Sure Unacceptable All animation clips and the choices in the task were ordered randomly. Distractors and fillers were unrelated to the study and were excluded from the analyses.

Bio-Data
Questions related to participants' backgrounds, such as age, gender, first language, other languages they speak, the type of instruction they received, etc. are attached to the questionnaire. Also, participants' consent was obtained by marking the bottom page of the bio-data.

Study Procedure
The administration of the task began by requesting participants to fill out the questions related to their background and sign the consent form. Participants were informed verbally as well as in the consent form that they have the right to withdraw from the study at any stage of administration. They were asked to answer all the choices by selecting one answer only from the scale for each choice. Each participant completed the task individually with the researcher. They watched the video clips one at a time and marked their judgments on the provided answer sheets. The researcher acted as a mentor to ensure that all participants did not miss to response to any choice, besides, to the grantee that participants would not go back to change their responses.

Data Analysis
IBM SPSS version 21 is used to analyze data after administration. Descriptive statistics of all experiment groups in the target items are computed. Based on the test of reliability, it appears that the test gives the value of .841. The test of normal distribution indicates that the p ˃ .05. Therefore, a parametric test is conducted for determining the significance between the variables. The t-test measures the significant difference between two independent variables whereas the paired t-test measures two related variables on a single test item. The motion event is analyzed based on: 'to' with the manner-of-motion verbs in the event with boundary-crossing, 'to' with manner-of-motion verbs without boundary-crossing, 'to' with path verb 'go' in the motion event with boundary-crossing. Also, it used manner as a complement and 'to' with path verb 'go' without boundary-crossing and the manner as a complement.

Results
The demographic characteristics of the sample presented that all participants were monolingual and females. The mean age of the advanced Saudi participants was 35.5 years, whereas, the mean age of the intermediate Saudi participants was 19.5 years. The mean age of native speakers was 45 years. All Saudi participants were classroom instructed for about nine years in the state schools. Table 1 presents the Acceptability Judgment Task results. All items are calculated based on the acceptable tokens supplied by participants. Table 1 presents the means and the standard deviations of the participants' acceptable judgments to the target items. It showed that Saudi participants at the two levels of proficiency accepted the co-occurrence of the directional preposition 'to' in the motion event with the boundary-crossing (Advanced Saudi speakers = 50 % -Intermediate Saudi speakers = 63 %). In contrast, native speakers supply low acceptable judgments (41 %) since the directional preposition 'into' should replace 'to'. No boundary-crossing is found, where all participants scored high acceptable judgments in this construction as it is the typical realization of their first languages. Both Saudi groups accepted to conflate manner with motion with 'to' in the event with and without boundary-crossing. Since this conflation pattern is the characteristic of the English language, native speakers supply high judgments in the event without  Vol. 10, No. 3; difference between the advanced Saudi speakers and the native group in the acquisition of 'to' with boundary-crossing. Such an insignificant difference is absent between the intermediate level Saudi speakers and the native group as it is pointed out in the above section. To summarize, the results support the claim that L2 speakers transfer the property of the L1 while acquiring L2 grammar at the onset of the development. However, the influence of the first language is resolved due to an increase in proficiency level and exposure to the input.
Concerning the result of 'to' and expressing manner as a complement, the second research question aims to examine if both groups of Saudi speakers accept to express the manner as a complement with the boundary-crossing event. The study findings reveal significant differences between the intermediate level of proficiency Saudi speakers and the native group and also between the advanced Saudi speakers and the group. This result is due to the manner as a complement is impossible in English. However, it is possible in Arabic when it involves boundary-crossing. This concludes that both Saudi groups accept expressing the manner as a complement. As a result, proficiency insignificantly effects acquiring this property because the statistics reveal no significant difference between the judgment of intermediate and advance Saudi groups. Similarly, a study by Kabli (2013) shows that the advanced and the intermediate Saudi speakers persistently accepted the manner as a complement in construction such as The lady went into the room running, unlike the English and German native speakers who rejected this construction. In her study, the descriptive statistics showed that the intermediate Saudi rated this construction as 65% while the advanced Saudi speakers scored 66%. In English and German where this construction is impossible, both German and English rated this construction as 25% and 32%, respectively.
The research seeks whether both Saudi groups accept the conflation of the manner with the motion as well as expressing the manner as a complement in the event involving boundary-crossing. The paired t-test reveals no significant difference in the performance of advanced Saudi speakers. They rated both constructions as an acceptable judgment of the event. However, advanced Saudi speakers started to recognize that 'to' is unacceptable with the boundary-crossing event due to an increase in proficiency, unlike the intermediate Saudi learners who accept this construction as 63%. Therefore, their judgments are lower than the intermediate Saudi speakers who are still constrained by their L1 in this stage of development as indicated by the preliminary result. Surprisingly, the performance of the intermediate Saudi speakers resembles that of the native speakers in the inferential statistics. The significant difference in both constructions stems from the fact that the native speakers generally accept the conflation of the manner with the motion regardless of the presence or absence of the boundary-crossing. Simultaneously, expressing manner as a complement is unavailable in English. As a result, a significant difference has risen. The question brought up is why the intermediate Saudi speakers' performance is dissimilar to that of the advanced Saudi speakers although both constructions are possible in Arabic. Upon interviewing some intermediate speakers later, Saudi speakers expressed difficulty to pronounce the sentence when manner as a complement in Arabic. For example, they find it easier to say in Arabic Sara ran to the room or Sara entered the room rather than Sara went to the room running. Therefore, they were more likely to accept the conflation of manner with motion than to express manner as a complement as is the result of the descriptive statistics revealed (63% vs. 39%). Slobin (2006) comes up with the concepts of heavier construction and the most common expression. He explains that the Satellite-frame languages usually prefer to select linguistically common expression. Therefore, he finds speakers of Dutch, German and English prefer to conflate manner with motion as the most common expression. 'In order to add manner to the perspective, [these] speakers … face the same processing problem as speakers of verb-framed languages: they would require a heavier construction, such as 'come flying out'' (Slobin, 2006, p. 10). Saudi students likely favor the conflation of manner with the motion more than expressing manner as a complement because of the influence of heavy construction in communication.
A plausible alternative account is addressed by McCawley (1978).
He introduces the concept of less linguistic effort in speakers' utterances. McCawley explains that the adjective pale can occur with many colors, such as pale green, pale blue, pale yellow. Colors, such as red, black, white are unlikely to co-occur with pale. The reason for lacking this combination is due to the existence of an alternative expression as mentioned by Householder (1971). He explains that the common expression of pale red in a single lexical item is pink. Thus, speakers prefer to use less linguistic effort when they say pink rather than pale red. In a similar vein, Levin and Rappaport Hovav state that 'Empirical studies show that information about manner is often omitted in descriptions of directed motion events with path verbs (Slobin, 1996b, pp. 212-213; see also Papafragou et al., 2006 on Greek). Expressions of the manner in PP and adverbial phrases are often considered heavy or unnatural (Talmy, 1973, p. 71), Levin and Rappaport Hovav (2015, p. 7). Since the conflation of the manner with the motion and the expression of the manner as a complement are available in Arabic, Saudi students acquire the habit to use the less linguistic effort of the conflation pattern in communication.
Intermediate Saudi speakers would likely change this habit in writing. Thus, this observation opens the doorway for further investigation in the future. the L1 influence continues at the advanced stage of development. The advanced Saudi speakers accept the same pattern of their L1 in acquiring manner argument structure. The outcome of the study is consistent with Stringer (2012) which supports that syntax of motion events for the L2 acquisition is associated with that of the lexicon.
The result shows a significant difference in the performance of all groups. The significant difference in the native group is a result of accepting the conflation of motion with the manner and rejecting the manner as a complement. Similarly, the Saudi groups accept the conflation of the manner with the motion and reject the manner as a complement because the event does not involve boundary-crossing as is the case in their L1. Also, this study proves that the element of crossing-boundary is an accurate test to distinguish languages within the typology, as proposed by Özçalişkan, (2013).
It may be worth noting that the limitation of this study lies in using one type of describing the boundary-crossing event in which the manner is expressed as a complement. It is better if the task included the other types of expressing boundary-crossing by using path verbs, such as enter, cross, arrive, ascend. By doing so, it would provide a solid background to test the degree of the preferred style of Saudi speakers in expressing boundary-crossing in the motion event. It hints that expressing the manner as a complement might be optional with path verbs. Therefore, it was desirable to add a choice that excludes the manner from path verbs construction to examine the performance of Saudi speakers in this form of construction. These limitations can shed light on further future research in the argument of the manner-of-motion to goal constructions.

Conclusion
This study investigates the influence of the L1 and proficiency in acquiring English manner-of-motion to goal argument structures. It answers the reason that Arabic encodes the manner with the motion in the main verb, in addition to expressing the manner as a complement. It supports that the L2 learners transfer the property of the L1 at an early stage of development. However, L2 learners show that they can overcome the L1 influence easily with an increase in proficiency, such as expressing 'to' with an event depicting boundary-crossing. Other constructions remain persistent in acquisition even with an increase of proficiency level, such as expressing the manner separately from the motion.
It has been acknowledged previously that the intermediate Saudi speakers acquired the habit to use the less linguistic effort of conflation patterns in communication. It is recommended for future studies to investigate if Saudi speakers maintain this habit in writing as well. Arab speakers usually use elaborate constructions in writing. Therefore, it is recommended to design a task based on describing the pictures in Arabic words. If it proves that Saudi speakers embrace this pattern of expression in writing, then it will be possible to say that the Arabic language has been changed over history, especially that the conflation form of less linguistic effort does exist in Arabic.