Revisiting Collocation: An Investigation Into Phraseological Units Muddle

Research on collocations still generates plenty of interest among contemporary researchers of linguistics and translation research. The interest stems from the fact that there is a degree of inconsistency in terms of identifying and defining collocations in the literature. Conflicting views put forward by linguists made it difficult to draw a coherent interpretation of collocations. This suggests that collocational units are a fragmented area of research with blurred lines between frequently fixed co-occurrence sequences or a combination of words and free word combination. The collocation debate has hit a conceptual impasse. It rarely moves beyond merely giving broad definitions, overlapped with other technical terms in the field of corpus linguistics. This study examines the extension to which collocations are intertwined with other formulaic expressions such as idioms and free word combinations. Furthermore, the study considers whether Quranic collocations fit into the general standards of collocation. The study examined a sample of five selected English translations of the Quran to evaluate their degree of faithfulness and accuracy. The findings revealed that collocation is too multifaceted to be pinned down to a single definition. Linguists fell short of reaching a consensus or providing concrete empirical evidence on the complex nature of collocation. The findings also showed that the English translation of the Quran remains a work in progress. Views are polarised between those who advocate a close rendering of the Quranic text and those who believe in a ‘natural style’. Translators need to be aware how collocations are formulated and how they are embedded in the Quranic verses to convey their deep and implicit meaning and should not be interpreted at face value.


INTRODUCTION
The term collocation comes from the Latin 'place together.' Collocations were first identified by Palmer (1933, cited in Nation, 2002 as a string of words that 'must or should be learned or is best or most conveniently learned as an integral whole or independent entity, rather than by the process of piecing together their component parts. 'Collocation; however, was first used in its linguistic sense by Firth (1957), who is credited with coining the term, and whose catchphrase 'You shall know a word by the company it keeps' is widely cited as a canonical reference in every study related to phraseology. This paper aims to explore the theoretical base of collocation and seeks to find out the extent to which Quranic collocational units demonstrate similar characteristics. It highlights the importance of exploring collocational and phraseological theoretical bases to facilitate the task of translators, academics and language teachers. It also intends to assess the difficulties and challenges of translating Quranic collocations from Arabic into English, focusing on five of the most referred to translations of the Quran in English. virtually no impossible collocations, but some are much more likely than others.' Take the classic example of 'blond'; the combination of terms 'blond horse' is unusual and unnatural. Thus, the meaning of collocation must not be viewed simply in terms of word associations but, according to Palmer (1984: 79), the meaning of collocation is 'idiosyncratic and cannot be predictable from the meaning of the associated words.' According to Van Der Meer, (1998:313) collocations are simply: 'words habitually grouped in clusters that are not considered idioms proper but are yet felt to be frequent and apparently belonging to the set of ready-to-hand units of language comprising more than one word.' Similarly, Benson (1990:23) argues, 'A collocation is an arbitrary and recurrent word combination. ' Likewise, Baker, (2018: 48) indicates that collocations tend to be 'largely arbitrary and independent of meaning and do not follow logically from the direct meaning of a word.' Moreover, researchers stress that collocations are often domain specific. The word file collocates with verbs such as create, delete, save when discussing computers, but not in other sublanguages (McKeown and Radev, 2000). Nation (2002: 318) stresses that all fluent and appropriate language use requires collocational knowledge.
Thus, meaning comes from word association and must be understood within the context of occurrence. The meaning of collocations is rarely straightforward and must not be taken at face value. For instance, fat chance may be rendered in Arabic as (unlikely) ‫حتمل‬ ُ ‫.غير.م‬ This also suggests that collocational meaning is not distributed identically in every language. An idea can be expressed differently in every culture. For instance, ‫ريعة‬ َ ‫أكوالت.س‬ َ ‫م‬ is translated into fast food instead of rapid/quick food, in English. In contrast, the French, refers to it as: restauration rapide (rapid food).
Collocational processing knowledge is important in understanding meaning, for as Harris (1968) claims, meanings of words are determined largely by their collocational patterns. Firth (1957: 11) stresses that, 'we best know the meaning of a word not by examining it in isolation but by the company it keeps. ' Firth (1957) seems to capture the essence of collocation and unites both academics and educationalists regarding the core meaning of collocation. For Choueka (1988: 67), collocations are lexical combinations that have a certain mutual expectancy. The lexical combination is not a fixed expression but there is a greater than chance likelihood that the words will co-occur. In contrast, Halliday (1966:153) views lexis as complementary to, but not part of, grammatical theory. He coined the notion 'set' as an extra dimension of the collocability of words, and a collocation in Halliday's definition is 'a linear co-occurrence. ' Halliday and Hasan later (1976: 320) introduced collocation as one of five general categories of cohesive devices for achieving lexical cohesion in the text: reference, ellipsis, substitution, lexical cohesion, and conjunction. Halliday and Hasan (1976) describe 'collocation' or 'collocational cohesion' as a cover term for the cohesion that results from association of lexical items that tend to occur in similar environments such as 'candle, flame, flicker ', and hair, comb, curl, wave'. The consensus is that collocating is a process of joining words together in phrases to form semantically unified expressions. Every word can be said to have a range of items with which it is more or less 'comfortable'. A range of items here refers to the set of collocates i.e. words which are associated with that particular word. As Van Roey (1990) states: [collocation is] that linguistic phenomenon whereby a given vocabulary item prefers the company of another item rather than its 'synonyms.' McKeown and Radev (2000) stress the difficulty in determining what acceptable collocation is, although it is clear that collocations occur frequently in similar contexts, which makes it observable. McKeown and Radev (2000:67) view collocations as 'those word pairs which occur frequently together in the same environment, but do not include lexical items which have a high overall frequency in language.' However, a solid theoretical base is still elusive because collocation represents an interface between language in use, and the diverse and creative shades of meaning which are being continuously generated, e.g. air miles, cyber cafés, blue sky-thinking, spin doctor, etc., remain challenging.
To conclude, it can be drawn from the range of views that no one has provided a holistic account of collocation dimensions, simply because this language phenomenon is at once too broad, slippery and lacks focus and consensus. Research on collocation so far, is yet to capture the full story about this complex aspect of language. In short, there is no one-to-one correspondence between two collocations across languages. A single collocation in one language may require two or more sentences to express the same meaning in another language, e.g. take a chill pill = cálmate! = calmez-vous! (Calm down).
Many of the definitions seem broad and overlapping in form and content, often giving the feeling of déjà vu, as having been repackaged and recycled from earlier definitions. Though formulated differently, most definitions share common themes. Some view collocation as a chunk of language, whilst others refer to collocation as a network that words build, metaphorically as a marriage contract between words, or the company that words keep.
There is a consensus regarding the marked and explicit features of collocation, and that collocation occurs frequently, which makes it perceptible. The key themes that emerged from the collocation debate are the same as those already known to define the meaning of collocation: • A combination of words that co-occur with predictable expectancy • A group of words which occur together frequently • Habitual co-occurrence of words • A close relation between words • Co-occurrence of two or more words • A sequence of words • A relationship a lexical item has with items • Arbitrary language recurrent in context • Word association • Word partnership One of the flaws that emerges from the collocation debate is that some collocations do not exhibit clear patterns and show some unpredictability and variability, the interpretation of which can be a cause of concern to translators. Herbst (1996: 390) for instance, views collocations, like idioms, as a combination of words that to some extent is neither transparent nor predictable. This view leads to the 'arbitrariness' side of collocation. It would be fair to say that collocation is too multifaceted to be pinned down to a single definition and to contend that collocation differs according to the language, the purpose and context in which it occurs. Thus, collocations in one language may be alien or untypical in another.

RESEARCH INTO COLLOCATION REMAINS INCONCLUSIVE
The collocation debate has generated a profusion of definitions, some of which have become so blurred and distorted that they have become ineffectual, defeating the purpose of creating a definition. The scale of lexical ambiguities and the woolliness of some definitions was illustrated earlier. This is not to denigrate research on collocation, which has produced some interesting insights. However, it is not free from criticism. For instance, Nesselhauf (2003: 224) contends that research on collocation is fragmented, inconsistent and 'unsatisfactory either because of the data-elicitation methods used or because of the vague definition of the concept of collocation.' Many collocation definitions contain loose and imprecise use of terminology. The challenges of getting a clear grasp of collocation should not be underestimated. For instance, Wray (2000: 465) identifies over 47 different terms related to collocation. She looked at collocation from a broader perspective in which she used the term "formulaic sequence" to cover a range of items "used to describe aspects of formulaicity in the literature" (2000: 464), defining it as prefabricated sequences that are stored and retrieved as a whole from the memory at the time of use.
To add to the confusion. Cowie (2009) uses different terminology and differentiates between chunks of language as set phrases and set sentences, whilst Nation (2002) uses the term collocation to describe both. Nation and Meara (2002: 36) view 'language units' as multi-word units. Furthermore, they use other different labels to refer to collocation such as 'pre-formulated language, formulas, and lexical phrases'. Nation (2002: 317) voices his concern saying that 'a major problem in the study of collocations is determining in a consistent way what should be classified as a collocation. ' Fontenelle (1994: 9) clearly stresses that: 'It should now be clear that there is no such thing as a clear, noncontroversial and all-embracing definition of a collocation. This notion should be conceived as a rather fuzzy area ranging from totally free combinations on the one hand to completely fixed multi-word units on the other. ' Thus, there is little agreement among researchers regarding the erratic and vague terminology as to what constitutes collocation. Further conflicting views emerge with regard to formulaic expression. It would be true to say that collocation theoretical debate has hit a conceptual impasse. This study argues that there are still many questions which remain unanswered concerning the nature of collocation and why some lexical items keep the company of one collocate rather than another, and what attracts one word to cohabit with another word more frequently than by chance.
The difficulty of identifying other key indicators that explain collocation characteristics other than habitual co-occurrence is another issue. Hasan later (1984) acknowledged that her original definition of collocation is too broad; therefore, she used the term 'lexical chain', which does not preclude confusion, as this new label is rather vague. In the same vein, Cowie (1994) argues that collocations are found in the 'fuzzy' area on a continuum between free combinations and idioms. Definitions tend to provide a general understanding of what collocation is. It covers all types of lexical co-occurrence. Within the umbrella term of the phraseological tendency; however, there are a number of terminological problems: both 'phraseology' and 'collocation' have been used in different ways and sometimes interchangeably. The multiplicity of terms has been seen as proof of inconsistency. This is further complicated by viewing collocations and idioms as two sides of the same coin or two ends of the same continuum, with collocations on one end, pure idioms on the other end, and figurative idioms in between (Cowie, 2009, Wray, 2008. Wray (2008: 10) adds to the conflicting debate by labelling the two ends as 'the contentious and the uncontentious,' with pure idioms existing on the latter end and collocations on the former. However, Nesselhauf (2003: 227) believes that the boundary between collocations and idioms is blurred: 'the line between collocations and idioms... is not rigid.' To sum up, the term collocation in its linguistic sense has since its inception by Firth in the 1950's, has given rise to prolific publications providing fresh insights, but it is probably fair to say that many authors fell short of reaching an accord or providing concrete empirical evidence on the complex nature of this word association. Conflicting positions and fuzziness still exist surrounding this linguistic phenomenon. Some position collocation in a continuum or as an extension with fixed phrases attributing to it, a hint of figurativeness; others contend that when pairings of words are expected and predictable, and where change or modification is not permissible except possibly in tense, it may be right to treat them as collocation.

DELINEATION OF BOUNDARIES
There is always a debate over which criteria are fit for purpose in deciding which combination/association/network of words is deemed to be a collocation. According to Bolinger (1979: 1) 'language does not expect us to build everything starting with lumber, nails, and blueprint, but rather it provides us with an incredibly large number of prefabs.' Research on collocation has so far provided neither a master plan nor a measurement by which collocation can be identified consistently according to clear benchmarks. No approach suggested is better than another or more appropriate for identifying collocation. Each has its own strengths and weaknesses, which actually varies depending upon the type of language and the nature of the text under consid-eration. Thus, a clear benchmark or pattern that precisely identifies collocation remains a contentious area with too many unanswered questions. The challenge stems from the fact that collocations are arbitrary and do not follow a prescriptive pattern or rules. Native speakers internalise them throughout the natural acquisition process; however, they do not come naturally for foreign language learners (Howarth, 1996;Baker, 2018;Cowie 2009, etc.) This study identified the broad parameters that put forward by researchers such as Jones and Sinclair (1974), Brezina et al (2015), Sinclair (2004), Baker et al. (2008) in order to corroborate the extent to which an association or a sequence of words is considered to be collocation: a) Frequency trend b) Phraseology trend a) Frequency trend. This view suggests that collocation can be identified as words that frequently co-occur together, as a kind of partnership. McKeown and Radev (2000) argued that collocations are those word pairs which occur frequently together in the same environment. This view; however, offers no semantic explanation for the type of relationship between the items of the collocational set. It is right to say that the way collocations are formed remains and continues to be subject to interpretation if arbitrariness, as a central element in shaping a collocational unit, is discarded. b) Phraseology trend. This refers to restricted collocations whose lexical items or clusters are fixed or idiom-like or as Cowie (2009: 67) puts it, have invariable opaque combinations and 'an element of figurativeness.' Thus, collocation has an embedded aspect of an 'idiomatic' element, e.g. hire and fire, no win no fee, armchair conscience. Moon (1997: 44) states that multi-word items have some degree of 'institutionalization, fixedness, and non-compositionality' which distinguishes them from 'other kinds of strings'. This suggests that collocations in addition to constraints of partnership, have some degree of figurativeness. Schmitt (2000: 77) also states that besides words co-occurring together, 'there must also be an element of exclusiveness. ' Schmitt (2000) goes on to argue that the fact that words in a collocation co-occur frequently implies that they are stored, and therefore retrieved when necessary as a single unit in the mind. This view is consistent with evidence from corpus analysis and psycholinguistics.

POSITIONING QURANIC COLLOCATION WITHIN THE COLLOCATION DEBATE
This section attempts to link the general theoretical base of collocation to that of the Quran in order to find out where Quranic collocation fits within the broad collocation debate. It also seeks to demonstrate whether Quranic collocations are in line with the widely accepted set of criteria for determining collocations. Although research on phraseological units is pervasively acknowledged as an established area in almost all language fields (Cowie, 1994;Römer & Schulze, 2009), there is paucity of collocational and phraseological studies between Classical Arabic and English. The language of the Quran is more formulaic than any other discourse (Bannister 2017). Quranic collocational and/or phraseological units are consistent and in line with the following collocation norms: • Frequent co-occurrence of two or more words • A close relationship a lexical item has with other items • Arbitrary language items recurrent in context • Word association, word partnership with a hint of figurativeness Lack of reliable and viable cursors for clearly determining the boundaries of phraseological units, across languages, is a well-acknowledged limitation. This semantic indeterminacy might lead to the conclusion of 'fence sitting' that some words may encode both literal and figurative meanings (Baker, 2018;Biber et al 1999;Cowie, 2009;Lakoff and Johnson, 1980). The formation of lexical collocations is not rule driven but arbitrary combinations of language items based on the assumption of frequency, high probability of co-occurrence and are context bound. Quranic collocation is viewed as one of the most powerful and persuasive means of expression which contributes to the ingeniousness of the Quran (Dweik & Abu Shakra, 2011;Ebrahimi, Pahlavannezhad, and Nadernezhad, 2012).This makes the task of identifying collocations based on an accurate set of criteria rather challenging. Research has found that a large proportion of the Quran is made up of collocational elements. The Quran encompasses a whole range of recurrent word combinations and many of the words form part of a frequent and habitual word-combination in one way or another. Some authors (Dweik & Abu Shakra, 2011;Zughoul & Abdul-Fattah, 2003) suggest that the Quran contains excessive collocational expressions in order to be persuasive, informative and influential.
The way collocation binds together with other language elements, and how different parts of the Quran are interconnected and explain each other, make any translation tentative. Every Quranic collocation has 'a cohesive force' and involves a unique encounter with language to achieve a specific communicative purpose. The collocation resonance and rhythm help in the memorisation of the Quranic verses. The way patterns of collocability are distributed between Arabic and English differ in terms of restrictions of positions and relationships, for example, the noun + adjective Quranic collocation ‫ريح.عقيم‬ reeh 'aqeem has an expressive metaphorical meaning. English uses different adjectives such as fierce, gale-force, high, strong, terrible to collocate with the noun wind. These adjectives; however, partially communicate the intended meaning of the Quranic collocation ‫ريح.عقيم‬ reeh 'aqeem. The rethorical purpose of the adjective ‫عقيم‬ 'aqeem remains unfulfilled in English translation.
One of the key themes that transpires from the above sample of collocation is that Quranic collocation features are consistent with the broad theoretical base on collocation, but they are also unique in the sense that they are used in a divine and creative way to convey a particular purpose. The way collocations are formulated is distinctive and the way meaning is distributed and emphasised may be untranslatable in English. Quranic collocational meaning is slippery, and there is a tendency to leave out significant information during the translation process; there is always more to say than the explicit meaning.

TRANSLATING QURANIC COLLOCATION
Research on how to translate collocation and in particular, the translation of Quranic collocation is limited to very sporadic and fragmented articles with little substance, which propose a number of different approaches and strategies to their translation. Translating Quranic collocation presents difficulties beyond those encountered in dealing with collocation in other languages owing to the Quran's style and complex structure. Translating Quranic collocation carries with it a warning, the risk of loss of meaning and vagueness and as such, it has been considered as one of the toughest challenges facing the translator. It has been treated as a part of the more general problem of untranslatability. Suffice to say that no translation is perfect; each has its strengths and weaknesses. The English versions of the Quran selected for the current study were carried out by Yusuf Ali (1034), Arberry (1955), Asad (1980), Khan & Al-Hilali (1983) and Abdel Haleem (2011. English translations of the Quran have been selected based on a number of considerations: 1. They are known among researchers for their relative accuracy. In other words, they are not known to include any deliberate deviations.

A preliminary comparison of these translations of a ran-
dom selection of Quranic collocations reveals that they use different ways of translating, which means the translators were independent and did not copy from each other. 3. The renditions selected for this study were done by translators of different tongues and cultural backgrounds. Some collocations are translatable, but many collocations cover certain areas of experience, which may be categorised, lexicalised and distributed in completely different ways by different languages. For instance, certain semantic combinations or associations made in Arabic may not be made by another language. To translate Quranic collocation efficiently, it is important to be aware and sensitive to the implicit shades of meaning that words may have. Take for example the collocate ‫َم‬ ‫ٍءا‬ ٍ . Like many collocations of the Quran, it fulfils the criteria of frequency, domain specific, transparency, close relationship with other items, arbitrary language items recurrent in context, word partnership with a hint of figurativeness etc. to the issue of quality of collocational units of the Quran in English, it could be seen as a matter of degree, i.e., it is difficult to measure. The debate over which is the best approach for translating Quranic collocation is still open for discussion and often generates conflicting views. The following illustrates how Quranic collocations were dealt with by five different translators.
There are discrepancies and variations in the way the collocate ٍ ‫اء‬ َ ‫م‬ 'water' was translated particularly in the example ٍ ‫ين‬ ِ ‫َّه‬ ‫.م‬ ‫َّاءٍ‬ ‫ّن.م‬ ‫ٍ.مِ‬ ‫ة‬ َ ‫ل‬ َ ‫ُل‬ ‫ن.س‬ ‫ُ.مِ‬ ‫ه‬ َ ‫ْل‬ ‫َس‬ ‫َ.ن‬ ‫ل‬ َ ‫ع‬ َ ‫.ج‬ َّ ‫م‬ ُ ‫.ث‬ Ali suggests 'quintessence of the nature of a fluid despised'. One of the reasons why this translation does not make much sense is that translating collocations depends on the way the collocation itself is perceived and understood. The closest meaning was produced by Hilali & Khan 'He made his offspring from semen'. Thus, translating collocation is not a simple matching between pairs of languages, but rather a product of the dynamic process of communication. This underscores that what forms a semantically correct meaning in one language may not be the same in another. The translator should be well aware of such formulaic expression complexities.
Quranic collocation is formed in a textual and contextual setting in which it aims to convey a specific message. The whole meaning of the collocational unit in the Quran is more than simply what the individual words actually mean -'an element of figurativeness' (Cowie, 2009: 67). According to Hoffmann (2007:33) 'Even prolonged cooperation between an expert team in the Arabic language, theology, philosophy, history, anthropology, psychology, sociology, literature, physics, and biology would never arrive at final conclusions.' English and Arabic users employ their respective languages from different grammatical systems and from different mindsets and from disparate thought processes; each operates from a different worldview so transferring Arabic collocation into English often leads to loss of meaning. (Amer and Menacere, 2013)  As can be seen, differences in the way the collocational patterns are formed and the way they are distributed in Arabic Quranic text can cause challenges to translators who may produce odd collocations in English. Collocation implicitly restricts the language user or translator to place randomly any word next to any other word. Many collocation researchers such as Harris (1968) and Firth (1957:11) contend that meanings of words are determined by their collocational patterns and words must be understood by the association they keep. Yet, it is fair to say that, on occasions, meaning is obtained from a wider area and beyond the scope of context.
It seems that translators experienced difficulties in rendering the collocate ‫َاسٌ‬ ‫ِب‬ ‫ل‬ libaas which implies something much more profound than the surface and literal meaning. The meaning of the term ‫َاسٌ‬ ‫ِب‬ ‫ل‬ libaas serves a specific communicative function and should have been extrapolated from various other sources, because translating is not mechanical. It is a process of interpreting, negotiating and accommodating the source language information so that it makes sense and it fits with the TL norms and usage. This agrees with Palmer's view (1984) that the meaning of collocation is distinctive and cannot be predicted from the meaning of the connected words.
It could be argued that the translators aimed at achieving accuracy at the expense of naturalness in T.L. The five translators decided to opt for a faithful translation, staying close to the original. Literally the term ‫َاسٌ‬ ‫ِب‬ ‫ل‬ libaas refers to 'an item of clothing', 'dress' but in this context this collocation has a much wider semantic scope. As for ِ ‫ُوع‬ ‫ج‬ ْ ‫.ال‬ َ ‫َاس‬ ‫ِب‬ ‫ل‬ libaasal joo'i was rendered 'garment of hunger' 'garment of famine.' This is a process of stringing word for word translation of the Quranic collocation in English, thus, creating a meaningless collocation. No word association or construction, or collocation of one language can have a matching equivalent in another. Meaning in this example needs to be inferred from other sources rather than directly transferred and taken at face value from the original.
In translating Quranic collocation, it may be necessary to use quite a different form to express meaning. There are different ways and forms of conveying the same idea. Hence, lack of comprehension of the source text often leads to stilted or distorted meaning in TL. In the above collocation ‫َاسٌ‬ ‫ِب‬ ‫ل‬ libaas was metaphorically used to refer to a man and woman in their relationship to each other, each protects the other for better or for worse ‫نَّ‬ ُ This study argues that the extent to which a collocation is alien or acceptable in T.L. depends on the overall effect it has on the receptors. The way information is conveyed differs from language to language.Quranic collocations are context sensitive; they serve a specific communicative purpose. Translators need to understand how collocations are formulated and how they are embedded in the Quranic verses to convey deep, implicit meaning and should not be interpreted at face value. It starts with grasping and unpacking the collocational word combination rather than taking across what the words or phrases might mean by themselves. For instance, consider the following collocate ‫شجرة‬ (tree) as it appeared in the Quran.
Moreover, the Quranic collocations ‫شجرة.الخلد‬ and ‫شجرة‬ ‫مباركة‬ can be translated literally as the tree of eternity or the tree of life or, as referred to in the Bible, as the tree of the knowledge of good and evil'. Similarly, ‫والشجرة.الملعونة.في‬ ‫القرآن‬ as the cursed tree. However, in reference to the collocation ‫,شجرة.الزقوم‬ the five translators did not try to explore in depth the meaning of ‫شجرة.الزقوم‬ Instead they kept the original as illustrated below: the Tree of Zaqqum? the Tree of Ez-zakkoum? the [hellish] tree of deadly fruit? the tree of Zaqqum the tree of Zaqqum (a horrible tree in Hell)?
The translators transferred the form of the word but failed to convey the spirit of the word which led to ambiguity and confusion, except probably Khan and Hilali who put between brackets the gist of what 'the tree of Zaqqum' means (a horrible tree in Hell). The best approach when dealing with Quranic collocation in translation is to consider the degree of translatability i.e. whether the equivalent expressions enjoy the same stylistic value in their respective contexts and cultures. This often involves adjusting or rephrasing as necessary to avoid mistranslation.

CONCLUSION
In closing, it can be claimed that research on collocation as a recurring word combination has a broad scope with the dual purposes of contributing to the theoretical knowledge base on collocation and to make collocation learnable and teachable to second language learners. Although collocation in its generic sense is understood to mean a close relationship that words form and then frequently appear together, it is often used as an umbrella term to refer to a wide variety of labels. Researchers fail to reach a consensus or provide concrete empirical evidence on the complex nature of collocational units. Conflicting perspectives and fuzziness still exist surrounding this linguistic phenomenon. The findings also show that translation of the Quran in English is still a work in progress. Views are divergent between those who advocate as close a rendering of the Quranic text as possible and those who believe in a 'natural style'. Literal translation appears to be the preferred method in translating Quranic