1. Introduction

A clitic is a grammatically unaccented bound morpheme that can be used in structures larger than words. It forms a phonological word together with its base lexeme. Phonological words are syntactic groups or structures that are phonologically pronounced like words but are not considered as words in terms of morphology, syntax, and semantics (Shaghaghi, 2009, p. 74). Clitics, like affixes, need a host to be attached to. However, unlike affixes, they are not considered as part of a word structure (Shaghaghi, 2009, p. 73). Clitics can be divided into four categories depending on their position in relation to the host word they are connected to: “proclitic” and “enclitic”, which are attached to the beginning or end of the host word, respectively. Furthermore, according to Harris (2002) and Kari (2003), there are “mesoclitics”, which appear between the host word and other affixes, and “endoclitics”, which split the base of the host word in two parts and are inserted between them (Jam, 2013, p. 62).

According to Castairs-McCarthy (1992, p. 142), clitics are syntactically similar to words and phonologically and morphologically similar to affixes. Zwicky (1994, p. 572) says that a clitic is an element that has similar properties to words but is not independent like them; it does not appear in a sentence by itself and is phonologically dependent on the neighboring word.

Shaghaghi (2012), in her “phrasal affixes”, analyzes the position of clitics in the Persian language structure. Accordingly, she considers clitics to belong to phrasal affixes or to the second-position ones. Those located at the edge of the group are phrasal affixes, while those that are not dependent on category of the base are second-position clitics.

Clitics are among the elements found in many Iranian languages and show interesting phonological, morphological, and syntactic features that need to be studied. Although much research has been conducted on these clitics, a deeper understanding of their form and function can be gained by looking at them from a cognitive perspective. The discussion of clitics has emerged in linguistic debates since 1970s. Clitics are placed somewhere between affixes and words. Some criteria have been proposed to distinguish them from the other two groups. Clitics are usually unaccented and depend on the host word to which they are attached; they appear on the final level of the word and no other affixes may be added afterwards; they are free to choose their host; they may go under syntactic movement and they are rule-governed. Clitics are divided into simple and special types. Simple clitics are reduced phonological forms of independent morphemes and usually have the same distribution as their free forms. Special clitics do not necessarily have the same phonological forms as their counterparts and there are certain rules for their distribution. One of the most important classes of clitics is the pronominal clitics which are found in various languages including the Iranian languages, and have attracted great attention of linguists due to their interesting features (Rasekh Mahand, 2013).

Simple clitics in Persian are reduced forms of the functional morphemes 'ra', 'va', and 'ham', which are commonly used in fast speech forms and casual conversations (Shaghaghi, 2014). This study investigates the frequency of simple clitics in 343 print sports media headlines. Though the language of sports newspapers is much closer to spoken language and colloquial speech than other newspapers, we should examine the rate of frequency of simple clitics in sports newspaper headlines and also their similarities and differences with spoken language in using simple clitics. For this purpose, within the framework of Shaghaghi's classification (1995), this type of clitics is studied and analyzed. Specifically, the questions of this study are as follows:

1. How frequent are simple clitics in sports newspaper headlines?
2. What are the similarities and differences between using simple clitics in sports newspaper headlines and spoken language?
3. Review of Literature

This section describes the most relevant previous research on clitics. It begins with studies conducted outside Iran on clitics including Jacob Wackernagel (1892), Spencer (1991), Zwicky (1977, 1985, 1994), Klavans (1985, 1982), Katamba (1993), Katamba and Stonham (2006), Gerlach (2002), Kruger (2005), Booij (2005), Katamba and Stoneham (2006), Gerlach (2002), Kroeger (2005), and Booij (2005) and continues with studies carried out in Iran including Shaghaghi (1995,  2007, 2010, 2013), Meshkat-o-Dini (2000), Kalbasi (2001), Rasekh Mahand (2005, 2010), and other Iranian studies. In Indo-European languages, clitics appear in second position after an accented element in a sentence (Wackernagel, 1892). 

Clitics are elements with some independent properties, but they are usually distinct from independent words; i.e. they cannot stand alone. They are phonologically attached to a host word. Clitics are usually grammatical words such as pronouns or auxiliary verbs. Historically, they are often derived from full words and eventually become inflected affixes. The set of phenomena known as cliticization is a meeting point of morphology, syntax, and phonology (Spencer, 1991).

Historically, the label “clitics” was originally applied by ancient Greek grammarians to elements that were syntactically words but were morphologically related to the neighboring words (specially the preceding words). Together with their base words, they form a kind of morphological unit (Zwicky, 1994). In describing clitics, Zwicky (1977) argues that in most languages there are morphemes that present analytical difficulties because they can be clearly classified neither as independent words nor as affixes (Zwicky 1977). He believes that taxonomic and theoretical assumptions about clitics require distinguishing them from words and inflectional affixes. It can therefore be found that the issue of distinguishing between words, affixes, and clitics is of great importance. A clitic is an element that has some of the properties of a full word but, unlike a word, lacks independence (Spencer, 1999). For this reason, Zwicky made great efforts to classify clitics. In this context, Zwicky (1977, 1985), Klavans (1982, 1985) and Zwicky and Pullum (1983) presented a set of tests and parameters that can be used to identify and classify clitics. Based on Zwicky and Pullum's tests (1983), bound morphemes can be divided into four groups: derivational affixes, inflectional affixes, simple clitics, and special clitics. The connection of a clitic with its base and its realm of function can be determined using the parameters proposed by Klavans. The first parameter determines the base or host of the clitic so that the clitic attaches to the first or last syntactic daughter constituent in the sentence and is called the initial or final constituent, respectively. The second parameter depends on the position of the clitic in relation to this syntactic daughter which may appear before or after that. The third parameter concerns the direction of the clitic attachment which is connected to the base as a proclitic or enclitic (Shaghaghi, 2012, cited in Klavans).

Katamba (1993) defines clitics as bound morphemes that can be attached to independent words by syntactically motivated rules. He calls these independent words as hosts. Clitics may appear at the beginning or at the end of their hosts, which are called proclitics or enclitics (Katamba, 1993).

Katamba and Stonham (2006), citing Zwicky (1997), view clitics as bound morphemes which occur almost in all world languages. Neither are they independent words, nor do they belong to affixes. They also state that, in addition to inflectional affixes, there is another class of bound morphemes which are called clitics. Clitics can attach to independent words using syntactically motivated rules. The word to which clitics are attached is called the host.

Some recent research suggests that clitics should be classified as a 'word' or 'affix'. Others propose the category of “phrasal affixes” to refer to clitics (Gerlach, 2002). As defined by Krueger (2005), clitics are linguistic forms that resemble both independent words and affixes. Booij (2005, pp. 168-166) considers clitics as “reduced words that refer to function words or non-lexical classes which depend on the host word used before or after them, such as pronouns and determiners, and they cannot be considered as phonological words by themselves” (Sarahi & Alinejad, 2013, p. 106). 

According to Iranian traditional grammarians, when linguistic elements such as “-am, -at, -ash, -eman, -etan, -eshan” function as an object, they are called bound object pronouns (Qarib et al., 1987, pp. 74-77; Natel Khanlari, 1980, pp. 198-199; Ahmadi Givi & Anvri, 1995, pp. 30-31). Later studies, however, have a different view on these elements and call them as clitics. Meshkat al-Dini (2000) believes that although they refer to a direct object (in terms of grammatical relations), they do not bear primary accent and are not part of the inflectional morphology; therefore, they are clitics. In Kalbasi's (2008, p. 25) view, a clitic is a kind of word that cannot stand alone and, like affixes, is attached to the preceding or following word; but, unlike affixes, it is not considered part of the structure of that word. Clitics that are attached to the end of a word are “enclitic” and clitics that are attached (added) to the beginning of a word are “proclitic”. Persian clitics are mostly enclitic.

Clitics are unaccented grammatical morphemes that do not appear in structures larger than words and phonologically form phonological words with their base. “A phonological word is a syntactic group or constituent that is phonologically pronounced like a word, but is not a word in terms of morphology, syntax, and semantics” (Shaghaghi, 2016, p. 74). In defining clitics, Shaghaghi (2010, p. 73) also writes, “in their researches in the second half of the 20th century, grammarians encountered dependent morphemes that played a grammatical role but were not included in the category of affixes. These morphemes were not involved in the production of words or word forms, but they participated in the field of syntax, i.e., in the construction of syntactic groups and sentences”. She adds, “on the one hand, clitics follow syntactic rules as free lexical and grammatical morphemes and words, but on the other hand, they behave like affixes because of their dependent nature; i.e. they necessarily attach to their bases. Unlike affixes, these elements are not considered part of the word structure and cannot be dropped”.

Analyzing the position of clitics in Persian syntax, Shaghaghi considers that clitics may belong to phrasal affixes or to the second-position. Those that attach at the margins of the group are group affixes, while those that appear free from base are second-position clitics (Shaghaghi, 2013).

Descriptive studies of clitics show that this element resembles words in some features and affixes in others. Zwicky (1994) notes that the categories of words and affixes are distinct in terms of syntax, morphology, semantics and phonology. Words have a special syntactic category and can potentially be the head of a group to which modifiers are added. Words are lexical morphemes and have independent stress; while affixes do not have a syntactic category and cannot be the head of a group; their meaning is based on syntactic relations and they attach to a morphological base. In the sentence /Gav ælæfha ra khord/ (The cow ate grass), the four morphemes "gav", "ælæf", "ra" and "khord" are examples of words, while the grammatical plural morpheme "-ha" is an example of an affix. Clitics, by their nature, have some of the features of words and some of the features of affixes. There is actually a continuum between words at one end and affixes at the other end. Clitics are divided into three categories: simple clitics, special clitics, and bound words. Reduced pronouns in English are simple clitics (e.g., in the sentence “give-m to me”, “-m” is the reduced form of “them”). Simple clitics share the same distribution with their full forms and they are clearly related to their full forms semantically and phonetically. For example, in Persian, the enclitic /-o/ is the reduced form of “ra”. The morpheme "ra" is pronounced as "/ro/" in colloquial speech and is reduced to "/-o/" when preceded by a word which ends in a consonant letter (Shaghaghi, 1995).

Klavans (1985) argues that it is possible to provide a single theory of clitics, but special distinctions must be taken into account in this regard. These distinctions are in the form of parameters that can indicate similarities and differences between clitic systems in different languages. On the other hand, she also emphasizes that different types of clitics in a given language may indicate different parameters. Klavans claims that what distinguishes clitics from words or affixes is that, unlike affixes, clitics (usually) attach syntactically to a particular group, and this group is considered as the realm of clitics function. Therefore, Klavans uses the term “phrasal affix” to refer to clitics in this context (Spencer, 1991, p. 377). Considering the parameters proposed by Klavans, it can be said that they all appear after the base and are enclitic according to the direction of connection. Each clitic is dominated by one or more groups.

Within the framework of Shaghaghi's classification (1995), which is the theoretical background of this study, she (1995) identifies three simple and six special clitics out of over a hundred bound morphemes analyzed. In all the nine cases mentioned above, clitics are unaccented and bound, which prevents the application of morphological affixation rule and there is no accidental gap in any of the cases. The distribution is similar to that of words and the ellipsis rule applies to all clitics except for the ezafe kasre and indefinite article '-i'; while the movement rule cannot be applied to any clitic. In other words, according to the accidental gap test, the distribution and ellipsis rules, clitics are similar to words; whereas considering the dependency test, affixation avoidance and movement rule, they resemble affixes, especially inflectional affixes. The dependency test also shows no similarity between clitics and words and (derivational or inflectional) affixes; thus, they exhibit an intermediate status between words and affixes.

Simple clitics in Persian are reduced form of the words 'ra', 'va' and 'ham', which are commonly used in fast speech forms and casual conversations. Special clitics can be divided into two categories: clitics that do not have a full form such as indefinite article '-i', indefinite determiner '-i', ezafe kasre, and emphasis marker '-ha'; and clitics that have a full form but do not have the same distribution such as the bound stems of the verb 'to be' and bound personal pronouns (Shaghaghi, 1995, p. 150). Therefore, the study and analysis of the use of simple clitics in the titles of sports print media are conducted based on Shaghaghi’s classification (2014). In this respect, the use of simple clitics is assumed and expected to be relatively higher than the use of them in other print media because the language variety used by this type of media is closer to the language commonly used in colloquial speech. It seems that clitics occur more frequently in this language variety, but it remains to be seen what the results of the study will be.

3. Method

In this study, we examine three functional morphemes which can be realized as simple clitics in print sports media headlines. This study is almost considered as a quantitative descriptive research study. In general, the data in this study was collected from a small corpus, that is, sports newspapers including Abrar-e Varzeshi, Piruzi, Esteghlal-e Javan, Khabar-e Varzeshi, Goal newspaper, Navad 90, and Iran-e Varzeshi. They were selected from May and June 2014; each newspaper office usually published 25 newspapers in May 2014 and 24 newspapers in June 2014. The method was based on selecting headlines of these newspapers, counting the number of simple clitics, and then comparing them.

In the analysis of every headline, functional morphemes including ra, va, and ham were observed and then calculated together with their clitic forms, which can be realized as simple clitics. In sum, a total of 343 headlines (175 in May and 168 in June) were selected and these grammatical morphemes, as the data for this study, were checked and counted in these headlines, then their frequency was considered comparing to each other. More importantly, the aim of this study was to examine if the simple Persian clitics are used in the headlines of sports newspapers as the print media and how much their frequency was.

4. Results and Discussion

In this section, we will examine simple clitics in print sports media headlines in the above newspapers. In practice, these data consists of simple clitics that are considered as bound morphemes, which have the same syntactic distribution as their full forms as well as clear semantic and phonetic associations with them (Shaghaghi, 1997).

The number of grammatical morphemes that can be used as simple clitics in the spoken language is calculated in the headlines and shown in the following tables.

Table 1

The Number of Grammatical Morphemes that Can be Used as Simple Clitics in Newspaper Headlines in May

Table 2

The Number of Grammatical Morphemes that Can be Used as Simple Clitics in Newspaper Headlines in June

Table 3

The Number of Grammatical Morphemes that Can be Used as Simple Clitics in Newspaper Headlines in May and June

Table 4

The Total Number of Headlines and Total Number of /ra/, /væ/, and /hæm/

Table 4 shows that there are only 74 words in about 343 headlines in the two-month period that can be used as clitics, but only in two headlines among all these cases, the simple clitic (bound) forms were used instead of its full form:

1- /Parvin: in pesare dige 'kardo' (kɑ:rd-o) be ostokhun resunde! (Goal newspaper, Wednesday, April 29, 2015). (Parvin: I am fed up with this guy!)

This is the first headline in which the clitic "o" instead of "ra" is used in the second position, after the noun in this noun phrase, and it is added to the base noun 'kɑ:rd-o ' (knife), which is used idiomatically rather than denoting a definite noun. Moreover, according to Halpern (1998), the string 'kɑ:rdo' is not necessarily a semantic structure but a phonological word that appears in the second position by attaching the clitic –ُ  (o) to the base noun.The second headline in which the clitic –o instead of "ra" is used is as follows:

2- /Amrica ra bezanid va halesho bebarid/ (Esteghlal, Thursday, June 18, 2015)

(Defeat US team and live it up.)

What is important in the headline “Amrika ra bezanid va halesho bebarid” is the use of two clitics ''-esh'' instead of " ɑ:n” (that)  and  “o” instead of '' ra'' : (halesho = hal-e ɑ:n ra). It should be added that in the pronunciation of hal-e sho, the clitic –e is also uttered, but in the prose/ writing, this –e is not written. Only –o is a simple clitic and –sh and –e are not simple clitics, but they are pronominal and special clitics, respectively. In another headline, “Perspolis! omideman ra naomid nakon”: (omideman = omid-e ma), "-eman/ -emɑ:n" here is a pronominal clitic which refers to "ma", while "ra" is independent. In this case, ''-eman" as a clitic is attached to the base, though it is not considered as a simple clitic and it is not the subject of this study. Generally, these clitics "-eman", ''-e","-esh", and "-o'' are enclitics which appear at the end of nouns, although they do not belong to the same category of clitics.

In other headlines (342 out of 343 ones), the words 'ra', 'va', and 'ham', which can be represented as clitics, are used in their full forms in the prose of these sports newspapers, a few examples of which are mentioned here:

3- /perspolis ba Branko ham nemi-barad/ (Abrar-e Varzeshi, Monday, April 27, 2015)

(Persepolis can't win with Branko, either.)

4- /Perspolis! Omideman ra naomid nakon/ (Piruzi, Sunday, April 26, 2015)

(Persepolis! Don't let us down)

5- /ham teractor ra mibarim ham Perspolis ra/ (Esteghlal-e Javan, Saturday, May 9, 2015)

(We’ll win both Tractor and Persepolis)

6- /Ladrop moch-e Branko ra khaband/ (Khabar-e Varzeshi, Thursday, April 23, 2015)

(Laudrup beat Branko)

7- /Branko: Daei va Karimi rofaghayam hastand. Dust daram biyayand sar-e tamrin/ (Goal newspaper, Wednesday, May 6, 2015)

(Branko: Daei and Karimi are my friends; I want them to attend the training sessions)

8- /nemitavanam ba pul-e khodam esteghlal ra bekharam/ (Navad 90, Tuesday, April 28, 2015)

(I cannot buy Esteghlal with my own money)

/Bazi-ye marg va zendegi/ (Iran-e Varzeshi, Wednesday, May 6, 2015)

(A game of life and death)

9- /esteghlal emruz ham malek peyda nemikonad/ (Abrar-e Varzeshi, Monday, May 25, 2015)

(Esteghlal can't find an owner today, either)

10- /Branko: bekhahand tim ra baraye qahremani mibandam/ (Piruzi, Thursday, May 28, 2015)

(Branko: If they want, I will arrange the team for championship)

11- /ta esteghlal hast, Hejazi ham khahad bud/ (Esteghlal, Saturday, May 23, 2015)

(Hijazi will exist as long as Esteghlal exists)

12- /Karimi: in aqayan va in football/ (Khabar-e Varzeshi, Tuesday, June 2, 2015)

(Karimi: these men and this soccer)

13- /Daei va Karimi ra ashti midahim/ (Goal newspaper, Tuesday, June 16, 2015)

(We’ll help Daei and Karimi be reconciled)

14- /shartbandi-ye yeki az nazdikan-e Daei Qahramani ra gereftar kard/ (Navad 90, Monday, June 1, 2015)

(A close relative of Daei put Qahramani into trouble by betting!)

15 /porfosor ham ruye hava/ (Iran-e Varzeshi, Monday, June 1, 2015)

(Professor up in the air, too!)

5. Conclusion

Simple clitics in Persian are the reduced forms of the words 'ra', 'va', and 'ham' and are usually used in fast and casual speech (Shaghaghi, 1995). Thus, we studied the use of simple clitics in the headlines of print sports media because the languages used in this type of media is closer to colloquial speech than other media. Therefore, it is expected that simple clitics are used more frequently in this type of media than others. In spite of this expectation, our study showed that in all cases where simple clitics might be used, their full forms were used except for two cases (kɑ:rd-o and  hal-e-sh-o), the dependent form "-o" is used instead of "ra". Totally, the data in the above sentences and their statistics in the tables show that among the words 'ra', 'va', and 'ham', 'ra' is the most frequent functional word, which was used 38 times; thus, it has been assumed that its clitic form might be more frequently used than 'and' with the occurrence of 26 times and 'ham' with the occurrence of 11 times. Only two occurrences of the simple clitic "-o" instead of ''ra'' were observed in the data. Thus, in answering the first question, it was found that the frequency of simple clitics in print sports media is very low and can be ignored. In other cases, even if the full form is used in writing, it is pronounced as its dependent form in reading. In most cases, this applies to the word "va" which is pronounced as "-o".

Simple clitics are basically used in spoken language not in written language, but sometimes they are used with very low frequency in sports newspapers and the prose of these sports media tends to be similar to colloquial speech. Furthermore, since the readership of sports newspapers comes from different walks of life (ranging from children, adolescents to adults), on the one hand, it seems that a fluent writing style which is close to the colloquial speech is required at first glance. On the other hand, the lack of higher number of simple clitics in sports newspaper headlines shows the fact that sports newspaper writing or prose is not very close to colloquial speech and has its own writing rules. Therefore, it can be concluded that, for fluency and correct understanding, the full forms of functional words "ra", "va" and, "ham" are used even in less formal sports newspapers comparing with other more formal newspapers like political ones, although the readers of which may be of a rather higher level of formality.

The results showed that, in spite of our expectation, due to the more fluency and proper comprehension of the text, simple clitics are rarely used in the prose of print sports media headlines and the frequency of the full forms of the functional morphemes of 'ra, va, and ham' is significantly higher than the frequency of their corresponding clitic forms.

From the theoretical point of view, simple clitics are fundamentally belonging to the spoken language, not the written one. In line with this linguistic fact, the findings show that headline writers of Persian sports media follow and observe this fact. Thus, they do not have a tendency to use simple clitics in their headlines, although utilizing this type of clitic is expected because the prose of these media is similar to the spoken language. Summing up, in answering the second question, it should be mentioned that there is a significant difference between using simple clitics in the print sports media and the spoken language.

The very limited use of simple clitics as the bound morphemes in headlines paves the way for the next researches to consider other types of grammatical morphemes in the newspapers in order to study their frequency and application in the prose of sports newspapers. In addition, the comparative studies on how to use clitics and other grammatical morphemes such as prepositions in sports headlines could be a research topic for future researchers.

Conflict of interest

The author(s) certify/certifies that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in the present research paper.