The comprehension of English compound nouns by Arabic-speaking EFL learners

Abstract

This paper examines the comprehension of English compound nouns by sixty Arabic-speaking EFL learners majoring in English at the University of Jordan, Amman. It explores the problems that these learners may encounter in understanding certain types of compounds, namely, subordinative, attributive and coordinative, based on the notions of endocentricity and exocentricity. Participants whose English proficiency level was advanced took a test to identify the meaning of an underlined compound, without using a dictionary. The responses to the three different types of compounds were analyzed using Two-Way repeated measures ANOVA. The results showed that the differences between endocentric and exocentric subordinative compounds were statistically significant in favor of the former. We argue that endocentric, especially subordinative endocentric compounds, were more easily understood due to their semantic transparency, i.e., the entire compound is a hyponym of its head, and to the similarity of this type of compound in the participants’ first language. The study provides some pedagogical implications for teaching compound nouns to EFL learners.

Keywords:

• Morphology
• compounding
• second language acquisition
• Arabic-speaking EFL learners

1. Introduction

Linguists have shown interest in compounding as a word formation process since it is common cross-linguistically. It is well known that compounding is one of the most productive processes in the morphology of many languages (Booij (2010: 93). Despite its productivity, scholars have not agreed on one general definition of compounds (Olsen, 2000; Lieber & Stekauer, 2009; Ralli, 2013 and others). For instance, Booij (2005) defined a compound as a combination of two words in which one word modifies the meaning of the other. Berman et al. (2009) stated that compounds are words composed of two (or more) bases, roots, or stems. Altakhaineh (2019) suggested that a compound is a complex word that consists of at least two adjacent words, where the non-head is normally non-referential. In spite of these differences in compounds definition, it can be observed that what makes a compound unique is the fact that the combination of two elements (whether words, stems, roots) yields a new meaning (A. Altakhaineh, 2022). Thus, it can be suggested that learning compounds could be challenging, especially for EFL learners. Since compound nouns are very common and new combinations are created almost daily, then it can be suggested that the mastery of compound nouns for EFL learners is crucial. Through examining the related literature, it seems that few studies have investigated the acquisition\learning of compound nouns by children and EFL learners belonging to different linguistic backgrounds (Chen et al., 2009; Nicoladis, 2002; Noorani & Salehi, 2019). Yet, studies that examined the comprehension of English compound nouns by Arabic-speaking EFL learners are non-existent, to the best of the researchers’ knowledge. Thus, this study aims to investigate the ability of Arabic-speaking EFL learners to produce the meaning of English compound nouns. Specifically, the participants were tested on three main types of compounds, i.e. subordinative, attributive and coordinative, each of which could either be endocentric or exocentric in relation to the way in which the components of the compound contribute to each other (Bisetto & Scalise, 2005). The following section provides the literature review of the study.

2. Literature review

2.1. English compounds

Within the framework of the semantic approach, it is argued that when two words are brought together, they produce a meaning which is more than the sum of two words independently. The resulting word is called a compound noun. According to Bauer (2001: 695), a “[c]ompound is a lexical unit made up of two or more elements, each of which can function as a lexeme independent of the other(s) in other contexts, and which shows some phonological and/or grammatical isolation from normal syntactic usage.” Thus, when the two elements of a compound are joined together, they generate a new meaning. As mentioned previously, the definition of compounds is rather elusive and various types of compounds can be identified based on several criteria, one of which is headedness. Based on the latter criterion, compounds can be either endocentric or exocentric. Endocentric compounds are the ones which have a semantic head (Booij, 2007). In endocentric compounds, generally, the compound as a whole is a hyponym of its head (Bauer et al., 2013), i.e., the head determines the overall meaning of the compound. For example, “bookshop”, which is a type of shop, has a semantic head which is the right element, shop. The second type, the exocentric compound, is not semantically headed. In exocentric compounds, the compound as a whole does not denote a hyponym of its head. For example, “egghead” is not a type of head; in this exocentric example, the overall meaning of the compound cannot be extracted from the meaning of its individual elements (Bauer, 1988).¹

Based on the grammatical relation between the elements of the compound, Bisetto and Scalise (2005) provided a classification of compounds also integrating endocentricity within. They proposed that this classification can be applicable cross-linguistically.² Syntactically speaking, this classification suggests that there are three different types of compounds, as shown in the following diagram:

The previous taxonomy shows two levels of classification, depending on the type of relationship between the elements of the compound. In the first level, there are three types—subordinate, attributive, and coordinate. The other level divides each compound into endocentric and exocentric compounds relating to semantic criteria, characterized by the presence or absence of the head as discussed earlier.

The two elements of subordinate compounds should have a complement relation; in these compounds, there is an element that functions as the complement of the head. Examples of endocentric subordinate compounds are taxi driver, coffee cup, and apple pie. The subordinate relation between the elements of the compound occurs even if the compound lacks a head, as in the exocentric subordinate compounds pickpocket, killjoy, and cut-throat. (Bisetto & Scalise, 2005). Melloni (2020) conducted a study on subordinate and synthetic compounds which are known to be widely affirmed modes of compounding across languages. These two types share some structural and interpretative similarities, but are still different in their classification parameters. While in subordinate compounds the emphasis is on the grammatical relation between the elements of the compound with a head-argument relation, synthetic compounds have their focus on the synthesis relation (morphological implications) between the compound members.

Attributive compounds are formed by a head as a second element and a non-head that modifies it and expresses one of its attributes, such as greenhouse, school bus, and windmill as which are endocentric attributive compounds, and egghead, red head, and red skin which are exocentric attributive compounds (Bisetto & Scalise, 2005).

In coordinate compounds, the first element of the compound does not modify the second; rather the two are equal, so in a sense, they are considered to have two heads. Examples of endocentric coordinate compounds are sweet-sour, blue-green, and doctor-patient, and examples of exocentric ones are north-east, mother-child and mind-brain (Bisetto & Scalise, 2005).

2.2. Previous studies on the acquisition of compounding and other formulaic expressions

Several studies (e.g., Charteris-Black, 1998; Cheng et al., 2011; Lioa & Chan, 2012; Azaz & Frank, 2017; Bormann et al., 2020; among others) have investigated the acquisition of compound nouns by EFL\ESL learners, but none of these studies dealt with the acquisition of English compound nouns by Jordanian EFL learners.

Charteris-Black (1998) investigated both the extent to which English compound nouns are problematic for Learners of English as a Second Language (ESL), and established factors influencing the comprehension of such expressions. The study was conducted with 34 students in the English for Academic Purposes (EAP) language support classes at the University of Surrey. It was designed to measure the comprehensibility of real (established) English compounds like blackmail, and invented (novel) ones like road rage. It was found that the idiomatic, syntactic factors and lexical novelty were factors behind such difficulty. The test included 27 novel compounds selected from the headlines of the Independent newspaper. The study motivated certain pedagogical implications based on the fact that the comprehension problems faced by L2 learners of English encountering compound neologism are very similar to those faced by native speakers.

Nicoladis (2002) studied the process of acquiring two languages, taking into account that a bilingual individual can more clearly illuminate the processes of language acquisition by children. The purpose of this paper was to examine cross linguistic transfer when acquiring compound nouns, that is, cross linguistic transfer would be clear if the compounds were reversed, e.g., toilet paper vs. paper toilet. The study was conducted with 25 monolingual English children and 25 French-English bilingual children between the ages of three and four years. They were asked about novel compound nouns by the researcher to evaluate their understanding of these new compounds. It was found that the bilingual children reversed compounds in English more often than the monolingual children when trying to produce the compound. While the subjects’ ability to comprehend the compounds was the same.

Tomoko (2009) emphasized the idea that, although collocations play an important role in learning a second language, especially for intermediate-advanced learners, the difficulties learners can face with collocations have not been investigated in detail.³ Verb-noun collocations like “take notes”, “place an order”, “cut corners”, and “make a discovery”, were examined regarding the students’ abilities to use them. In addition, Tomoko tried to show the importance of the instructions in guiding the learner to the input and restricting the combinations that they could produce. The participants in the study were sixty Japanese students, including 30 at intermediate and 30 at advanced level of English proficiency. First, they were tested using a fill-in-the-blank pre-test. Next, an instruction session was provided to brief the learners on the concept of collocations, discuss common mistakes in relation to collocations, and highlight the different restrictions on collocation between English and Japanese. After explaining the concept of collocation for the subjects and providing them with enough examples using flashcards, they were given paper-based exercises on collocations to study at home. After two weeks, the participants came back for a fill-in-the blank post-test. In the tests, errors were found such as: inappropriate paraphrasing, misusing certain verbs, blending two collocations with similar meanings, using morphological synonymy incorrectly, inserting unnecessary articles, particles, and prepositions between verbs and nouns, mixing intransitive and transitive verbs, forming collocations out of compound nouns, and some phonological errors. These reflected the difficulties the learners faced when trying to learn the collocations of another language, as well as the impact of their native Japanese on the second language English. The acquisition of compound nouns was also examined from another perspective. i.e. through applying a specific teaching method and exploring whether such a method can have an impact on the acquisition of compounds. For instance, Lioa and Chan (2012) conducted a study based on the idea of learning through play, in which a vocabulary game was used as a tool for learning compound nouns. The participants in the study were 79 low-motivated Taiwanese college students whose level of proficiency was A1 (pre-basic) and A2 (basic) at the CEFR placement test. The games that were utilized as a teaching tool were poker, chess, and Gobang, which were enhanced by several memorization techniques. The study tried to determine the effects of the invented games on vocabulary learning across subgroups of different proficiency levels. To answer this question, the data of the study were collected using 174 compound-nouns as a material of the pre-test, after-poker test, after-chess test, after Gobang and post-test. The collected data were analyzed using ANOVAs measures and T-tests to describe the statistics and samples. The results showed that the vocabulary games were important in facilitating the learners’ lexical growth in relation to compound nouns. This study may suggest that due to the importance of compound nouns, teachers utilize various teaching methods to help EFL learners acquire them.

Cheng et al. (2011) investigated the processing of compound nouns by four groups of Chinese/English bilingual children. They were placed into groups based on their level of proficiency in the two languages. The researchers designed a task that included compound words from the two languages. The task also included two types of compounds: transparent compounds (endocentric) such as toothbrush, and opaque compounds (exocentric) such as deadline. Four main results were found which give insight into the ways in which compounds are decomposed in English. First, the transparency of the compounds affected the speed of their acquisition. Second, the degree of lexicality of the translated compound nouns in the non-target language (Chinese) affected the accuracy of the response in the target language (English). Third, the lexicality of the translated compounds had no effect on the semantic transparency of the compound. Fourth, the response patterns were similar across all the groups, regardless of the level of language proficiency, i.e., language proficiency had no effect on the ability to respond using the compounds.

Azaz and Frank (2017) investigated the role of high/low perceptual salience in the acquisition of Arabic construct state by native English speakers. The study investigated whether contrasting patterns in head direction between Arabic and English helped or hindered the early stage of the definite construct state acquisition, and whether the degree of salience affected the sequence of definite construct state acquisition by English-speaking learners of Arabic. Forty-five English individuals were divided into three groups: beginner, intermediate, and advanced learners. Additionally, ten native Arabic speakers were used as a control group to test the validity of the test. The participants were asked to fill in the blanks of each sentence by looking at two accompanying pictures which helped the participants decide on the missing items. The responses were examined in relation to the head direction and position of the definite marker. Responses in which the direction of the head were correct and the definite article was added to the complement received a value of one. However, if any part was missing, the response received a value of zero. Incomplete responses were excluded from the analysis. The findings helped establish a number of pedagogical implications and raised questions that need to be answered in the future. The study concluded that the higher the salience, the faster the acquisition. Accordingly, the native English speakers were able to master the head direction phase before the definiteness phase since it is higher in salience. Furthermore, it was found that L1 effects in the acquisition of head direction, and any differences in the direction of the head, hindered the early L2 acquisition of such structures (Azaz and Frank, ibid).

Zibin (2016) explored the ability of Jordanian learners of English as a foreign language (EFL) to comprehend formulaic metaphorical expressions, and investigated the role that the first language played in the comprehension of such expressions. The study discussed six types of metaphorical expressions to: measure Jordanian EFL learners’ ability to comprehend metaphors in English; access L1 metaphorical knowledge in the comprehension of L2 metaphorical expressions; and, identify the problems that the EFL learners might face in comprehending such expressions. A 24-items multiple choice test was used to test the participants’ understanding of English metaphors. It was found that the participants found those expressions which are similar in both Arabic and English at the conceptual and linguistic levels easier than those which are different in the two language. In addition, the participants encountered difficulty with semantically non-transparent metaphorical expressions. Zibin (2016) introduced pedagogical implications to help EFL learners acquire English metaphorical expressions.

Bormann et al. (2020) discussed the role that verbal short-term memory (vSTM) plays in learning novel compound nouns. The study investigated a 61-year-old speaker of German with impaired vSTM. It was found that such impairment affected the process of word learning and acquisition of novel noun-noun compounds. Two independent experiments reflected the participant’s impaired learning of nonsense compounds, such as ball door and the names of unknown tools such as a nail puller. The study concluded with the fact that the vSTM is involved in the acquisition process of compound nouns. If it is impaired, the ability to establish links between elements in the lexicon is impaired since it connects phonemes to generate novel word forms.

The review of related literature demonstrates that compound nouns are regarded as important for EFL learners to acquire due to their frequency and productivity in English; thus, it remains an area worthy of further examination. The review also reveals that the acquisition of English compound nouns by Arabic-speaking EFL learners has not been explored to the best of the researchers’ knowledge. Hence, this research study aims to bridge this gap and to provide some pedagogic implications for teaching compound nouns to Arabic-speaking EFL learners.

3. Research questions

The study aims to answer to the following questions:

1) To what extent do Arabic-speaking EFL learners comprehend compound nouns in English?

2) Do Arabic-speaking EFL learners encounter problems in comprehending certain types of compounds compared to others? If they do, what are these types and why are they particularly difficult?

4. Methodology

4.1. Methodological approach

As mentioned previously, the study mainly aimed to investigate Arabic-speaking EFL learners’ ability to comprehend three types of compound nouns. To this end, a comprehension test was used in which the participants were asked to provide the meaning of 18 compounds. Below is a description of the sample, data collection tool, and procedure.

4.2. Sample

The participants of the study were sixty students majoring in English at the University of Jordan. Their mean age was 18. The participants have studied English as a compulsory subject at school for 12 years; however, none of them were native speakers of English. The Oxford placement test was employed to determine the English proficiency level of the first-year students who took part in the study. The test showed that these students had almost the same level of proficiency, i.e. advanced. The original number of participants was 80 but 20 students lacked sufficient proficiency, based on the results of the placement test, to produce the meaning of compound nouns and were excluded. The sampling procedure was convenient sampling since the researchers had access to the participants and the latter were readily approachable which facilitated data collection (see, Altakhaineh et al. 2022). With regard to research ethics, the researchers obtained consent from all participants to ensure that they know every aspect of their participation. That is, only willing participants took part in this test. All the participants were only instructed to write their gender, while their names remained anonymous; thus, their identifying information were not known to the researchers. The researchers also informed them that this test has no effect on their overall mark in the course; it was only a diagnostic test. Table 1 provides the distribution of the participants.

Table 1. Distribution of the participants

Gender	Number of participants	Score in the placement test
Female	49	120–140
Male	11	120–140

4.3. Data collection

The number of compound nouns selected for this study, i.e. Eighteen, is similar to that used in studies found in the relevant literature (Altakhaineh, 2017: 104–109; Scalise & Bisetto, 2009: 45–47; Berman et al., 2009: 9–11). The selected English compounds were then checked with reference to a corpus to ascertain their occurrence and frequency in contemporary speech. The selected corpus was the Corpus of Contemporary American English (COCA, 2022) which contains more than one billion words of text. Table 2 shows the items used in the test. To ensure the validity of the test, a pilot version was given to ten native speakers of English before the test took place.

Table 2. The items of the test and their frequency

Types of compounds	subordinate	frequency	attributive	frequency	coordinate	frequency
endocentricity	apple pie	1432	swordfish	823	blue-green	954
	taxi driver	1288	windmill	1079	doctor-patient	28
	coffee cup	1087	greenhouse	209	sweet-sour	54
exocentricity	cut-throat	128	red head	1669	north-east	273
	pickpocket	255	egg head	356	mind-brain	11
	Killjoy	153	red skin	80	mother-child	8

4.4. Instrument and procedure

As mentioned before, the participants undertook a test to identify their English proficiency level. The Oxford Placement Test was used since this examination is used by many education authorities in many countries as a measuring tool for the required standards in English.⁴ Specifically, the test was bought online and printed out to be administered in a course that targets first year students (n = 80) at the University of Jordan. One of the researchers administered the diagnostic test and marked the papers based on the guidelines provided. Participants were regarded as advanced learners if they score 120–140.

After excluding the 20 low-level participants, the remaining 60 took a comprehension test. Specifically, an 18-item test (see Appendix 1) with the question “What does the underlined compound in the following sentence mean?” was introduced to the participants to test their ability to identify the meaning of the underlined items without using a dictionary. The sentences listed in the test were extracted from previous studies conducted on compounding (Altakhaineh, 2017: 104–109; Scalise & Bisetto, 2009: 45–47; Berman et al., 2009: 9–11). The answer was considered accurate if it gave the correct meaning of the underlined compound. For example, for “My teacher is an egghead”, if the answer delivered the intended meaning of the underlined compound as smart, clever, well-educated, or intellectual, it was considered correct. The test was conducted to test the students’ competence of three types of compound nouns. These compounds were discussed in the theoretical framework section of this paper.

4.5. Statistical analysis

The differences in the scores of the test regarding the three different types of compounds were established using Two-Way repeated-measures ANOVA or Two-Factor repeated measures ANOVA (see, Potvin & Schutz, 2000). This statistical test was chosen since it makes it possible to notice variation between the scores of the different compound nouns, i.e., whether there are statistically significant differences between the results (Charteris-Black, 2002; Zibin, 2016).Specifically, this test compares means across 1 or more variables based on repeated observations, so it contains at least 1 dependent variable (i.e. the participants scores) that has two observations or more (i.e. three types of compounds and two types of each within). The test was used to determine whether the differences between the participants’ answers on the three types of compounds are statistically significant, taking into account that each type is divided into two types, namely, endocentric and exocentric compounds. The results of the two-factor repeated-measures ANOVA and their interpretations are discussed in the following section.

5. Results

This section introduces the results of the test conducted for this study. The distribution of correct answers is presented in Table 3 below. As can be noticed, each test item received a different number of correct answers.

Table 3. Distribution of correct answers

Types of compounds	Endo/Exo-centricity	Items	Number of correct answers	Total	Total
Subordinate	Endocentricity	apple pie	50	109	141
		taxi driver	32
		coffee cup	27
	Exocentricity	cut-throat	0	32
		pickpocket	14
		killjoy	18
Attributive	Endocentricity	swordfish	20	67	95
		windmill	23
		greenhouse	24
	Exocentricity	red head	11	28
		egg head	12
		red skin	5
Coordinate	Endocentricity	blue-green	14	57	79
		doctor-patient	16
		sweet-sour	27
	Exocentricity	north-east	8	22
		mind-brain	8
		mother-child	6
Total					315

Table 3 shows that the compounds are classified into three types: subordinate, attributive, and coordinate, according to the grammatical relations between the compound’s constituents. The grammatical relation between the parts of the subordinate compounds is a relationship of complement, i.e., the first component is the complement of the second component, as in “apple pie”, where apple is the complement of pie. With attributives, the relationship between the elements is of modification since one element modifies the other, as in “red head”, where red reflects one of the properties of the head. The constituents of coordinate compounds are linked together by a conjunctive relation (Mortensen, 2003) such as the relationship between the two elements in the compound “mother-child” which means of, relating to, or designating the relationship between a mother and her child.

To answer the first research question which is concerned with the extent to which Arabic-speaking EFL learners comprehend compound nouns in English, the results show that the total number of correct answers scored by the participants is 315, so compared to the total number of answers that should have been scored (60 (number of participants) × number of items on the test (18) = 1080), the percentage of correct answers is 29.2%. This result suggests that the participants have little knowledge of English compound nouns.

To answer the second research question which is concerned with the difficulties the participants encounter in comprehending certain types of compounds compared to others. Table 3 reveals that the participants’ accurate scores on endocentric compounds were higher than their score on exocentric compounds in all three types of compounds. That is, the participants’ accurate answer on subordinative endocentric compounds was 109 compared to 32 on exocentric compounds. Their answers on attributive endocentric compounds were 67 compared to 28 on exocentric ones, and 57 on coordinate endocentric compounds in comparison to 22 on coordinate exocentric ones. Since the exocentric compounds do not denote a hyponym of their heads, their meaning is not semantically transparent and thus are not as easily decipherable as that of endocentric compounds which are semantically headed (see, Zibin & Altakhaineh, 2018). Moreover, the class exocentric compounds denote is not derived from its constituents.

With regard to the difficulties the participants encountered with the three types of compounds, Table 3 shows that the number of accurate answers on the subordinative compounds was the highest (n = 141) followed by attributive compounds (n = 95) and finally coordinate compounds (n = 79). The subordinate compounds, especially endocentric compounds, were the easiest to decipher since the relation between their constituents is one of complementation, in which one of the elements is the argument of the other, e.g., Taxi driver is someone who drives a taxi. Such a relation is described as head-argument dependency (Melloni, 2020). However, on viewing Table 3, it is clear that exocentric subordinative compounds were harder to comprehend due to the lack of semantic compositionality between the elements of the compounds. For example, there is no element in the compound “cut-throat” that suggests the meaning of the overall compound as ruthless (see, A. Altakhaineh, 2022 for more details).

Attributive compounds were more challenging compared to subordinative compounds since the first element can express several relationships with the head, e.g., wind mill could be a mill that is operated by wind or a mill that produces wind. This suggest that the meaning of attributive compounds is not as straight forward as that of subordinative ones. Attributive endocentric compounds are easier to comprehend compared to exocentric ones since the former is a hyponym of its head, i.e., the element on the right can reflect the meaning of the whole compound, especially in case of endocentric attributive compounds. For example, “swordfish” is a fish and “greenhouse” is a house. The exocentric attributive compounds are also referred to as bahuvrihi. According to Bauer et al. (2013: 465), in bahuvrihi compounds there is a person or thing that has X, in which X is the property described by the compound. In “red head”, “egg head”, and “red skin”, the semantic head, i.e. person, is absent. Redhead is used to refer to a person who has a reddish hair. “Egghead” refers to a person who has a huge amount of knowledge, i.e., an intellectual. “Red skin” is used when describing a Native-American. Thus, the meaning of the exocentric attributive (bahuvrihi) is more challenging to comprehend compared to attributive endocentric compounds.

Finally, Table 3 also shows that coordinate compounds were the hardest to comprehend as they are elaborate expressions that have parallel structures. Bisetto and Scalise (2005: 327) suggested that such compounds can either be two-headed or headless, so deciphering their meaning is hard. Furthermore, the order of constituents within a compound cannot be reversed without changing the meaning. This detail is not clear for the majority of the EFL learners and can make the process of comprehending such compounds much harder.

To test whether the differences between the three types of compounds and the two type of each within are statistically significant, Table 4 presents the results of the two-factor repeated measures ANOVA.

Table 4. Two-way repeated measures ANOVA

Source of Variation	Sum of Squares (SS)	DF	Mean Squares (MA)	F	p-value
Subordinative (A)	1266.7	1	1266.7	24.9464	0.0003
Attributive (B)	345.3	2	172.6	3.4004	0.0676
AB	179.1	2	89.5	1.7637	0.2131
Error (residual)	609.3	12	50.7
Total	2400.5	17

Metaphorical expressions with non-transparent compounds so that they can understand the differences between endocentric.

Table 4 shows that the interaction occurred within the two types of compounds, namely, subordinative and attributive (there is no interaction with coordinate). However, despite the participants’ different scores on the three types, there were no statistically significant differences between them. In particular, F(2,2) at 0.05 level of significance = 3.8853, but as calculated FC = 3.4004 < 3.8853, suggesting that there were no statistically significant difference between the three types of compounds. On the other hand, there was an interaction within the two types (i.e. within in subordinative compounds, that is, the differences between the participants’ scores on endocentric and exocentric subordinative compounds were statistically significant (i.e. p= 0.0003 < 0.05), suggesting that the participants found endocentric subordinative compounds to be easier to comprehend in comparison to subordinative exocentric compounds. However, despite the different scores between endocentric and exocentric compounds within the two types, attributive and coordinative compounds, the differences were insufficient to be statistically significant.

6. Discussion

Drawing on the results discussed in the previous section, it appears that the participants found the processing of certain types of compounds more challenging than others, i.e., exocentric compounds were much harder to process than endocentric compounds. This result is similar to Cheng et al.’s study (Cheng et al., 2011), in which it was stated that transparent compounds (endocentric compounds) are easier to acquire than opaque compounds (exocentric compounds). The results also support those of Zibin (2016), who suggested that non-transparent formulaic expressions are more difficult to comprehend compared to transparent ones. By taking a closer look at the subordinate compounds, it can be noticed that the endocentric compounds were easier to comprehend than the exocentric compounds. When the head, the hypernym or the superordinate exists, the entire compound denotes a hyponym of this head making the meaning of the compound rather clear (i.e. apple pie is a hyponym of pie). This suggests that the subordinate, attributive, and coordinate endocentric compounds were easier to comprehend than the subordinate, attributive, and coordinate exocentric compounds although there were no statistically significant differences between them expect for endocentric and exocentric subordinative compounds. On the other hand, exocentric coordinate compounds such as “north-east”, “mind-brain,” and “mother-child” were the most challenging to process. This argument is also supported by the participants’ answers on the attributive exocentric compounds or bahuvrihis in which the head, i.e. person is missing and exocentric coordinate.

In sum, it can be suggested that three factors may affect the comprehension of English compound nouns by Arabic speaking EFL learners: the participants’ familiarity with compounds (cf., Zibin, 2016), the notion of headedness (the existence of a semantic head or its lack thereof; Zibin & Altakhaineh, 2018) and the influence of the learners’ first language. For example, some English compounds are same in Arabic, namely, subordinative endocentric compounds; thus, the participants can easily guess their meaning, e.g., coffee cup. However, differences between English and Arabic can affect this comprehension. For instance, the direction of the head in Arabic is different from that in English. In the former, the head is the leftmost element of the compound, whereas in the latter the head is rightmost element of the compound. Thus, in the other types of endocentric compounds, namely, attributive and coordinate, the participants’ correct answers were less compared to subordinative endocentric compounds. The next section provides more detail on these two factors and some pedagogical implications.

7. Pedagogical implications

Based on the results, it can be argued that Arabic-speaking EFL learners face obstacles in comprehending some types of compounds in English, despite the fact that compound nouns are part of their native language. These results cannot be generalized since the sample only included advanced learners; hence, students that have different levels of English proficiency may encounter other problems with English compound nouns. Additionally, it can been suggested that L1 possibly plays a role in the comprehension process of L2 compounds, especially subordinative endocentric compounds where positive transfer can take place, that is, the participants could have used literal translation to produce the meaning of the compounds apple pie, taxi driver and coffee cup since these compounds exist in Arabic, i.e. ke:ket ttuffa:ħ “apple cake\pie”, swwag ttaxi “taxi driver” and findʒa:n lgahweh “coffee cup”. These compounds are also quite common and the participants may have encountered them before either at school or at the university. Therefore, prior to explaining compounding in the L2, teachers are advised to present subordinative endocentric compounds in the students’ native language to help them understand their English counterparts (cf., Zibin et al., 2020). Once teachers are satisfied with the students’ understanding of compounding in their L1, they can move to compounding in L2 and discuss the similarities and differences between the two languages. The teachers also need to explain the differences between L1 and L2 with regard to compounds, e.g., the direction of the head. For instance, students should not be encouraged to use their L1 to decipher the meaning of other types of compounds especially exocentric ones because there are no similarities between the two languages in that regard. That is, the aspects which could be similar between the two languages as far as compound nouns are concerned should be first explained, then the differences should be illustrated. Based on our experience as EFL instructors, it is suggested that the use of compound nouns should be integrated more into the L2 curriculum at the school and university levels. Furthermore, teachers are also advised to illustrate the differences between subordinate, attributive, and coordinate compounds and between endocentric and exocentric compounds by illustrating the relations between the elements of the compound. In particular, the notion of semantic head in various constructs should be explained explicitly to EFL learners so that they can understand the differences between endocentric and exocentric compounds. Introducing pedagogical implications to help in the learning and teaching of compounds is similar to what was suggested by Zibin (2016), who introduced some pedagogical implications that can help EFL learners acquire English metaphorical expressions.

8. Conclusion

The main conclusions are: subordinative compounds are easier to comprehend by Arabic-speaking EFL learners compared to attributive and coordinate compounds, subordinative endocentric compounds were the most semantically transparent ones for the participants, and endocentric compounds across all three types are easier to comprehend than exocentric compounds. It was proposed that the notion of a semantic head plays a role in the participants’ ability to comprehend the compounds, i.e. the presence of a semantic head makes the meaning of endocentric compounds more transparent and in turn easier to decipher by the EFL learners. The participants’ familiarity with English compounds and their first language (L1) can also impact their comprehension of these compounds especially if their L1 is different from English. The study concluded with pedagogical implications that may help teachers in teaching compounds to EFL learners. In line with Zibin et al. (2020), studies which investigate the differences between Arabic and English formulaic expressions such as compounds are needed to shed light on the factors affecting the acquisition of these expressions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors received no direct funding for this research.

Notes

1. Creative compounds which include metaphor and\or metonymy subsumes both endocentric and exocentric compounds as argued by some researchers, e.g., Benczes (2006) and Zibin and Altakhaineh (2018).

2. It should be noted that another classification was proposed by Scalise and Bisetto (2009) in which another level was added to show the different types of semantic/interpretive relations between the two elements of attributive and subordinative compounds, but not coordinate. Thus, we chose to follow their 2005 classification as we could not find enough examples on each category based on the 2009 classification; we needed three examples of each category.

3. Compound nouns, as discussed previously, are defined as two or more words joined together to create a new meaning which functions as a single unit, e.g., bookshop, whilst collocations are defined as words or phrases that are commonly used together (with a frequency that is greater than chance), e.g., deep sleep.

4. https://elt.oup.com/feature/global/oxford-online-placement/?cc = jo&selLanguage = en.

Appendix

The Test

IELTS Score:

Instruction: Provide the English meaning of the underlined compound noun:

1. My teacher is really an egghead.

2. In my farm, I built a greenhouse.

3. My favorite dessert is the apple pie.

4. The mother-child bond is the most beautiful bond.

5. My company has fallen victim to cut-throat competition.

6. I saw a pickpocket in the supermarket.

7. I don’t like to be around him because he is a killjoy.

8. Swordfish live throughout much of the world’s oceans

9. I like sweet-sour food.

10. My friend is a redskin.

11. Trump is a redhead.

12. My dad is a taxi driver.

13. I like blue-green T-shirts.

14. The doctor-patient relationship of trust is a must.

15. The factory pumps the water using windmills.

16. The north-east are two directions found on the campus.

17. Doctors who study the human body know the difference between the mind-brain.

18. This is a special coffee cup. It is made of glass.