The Persian light verb dādan ‘to give’: Causation and more

1. Introduction

There are relatively a limited number of simple verbs that are commonly used to form complex predicates in Persian¹ This usage has persuaded some linguists to argue that Persian is a sterile language (see for example, Bateni, 1989). This means that Persian is endowed with very few simple verbs. Therefore, in order to generate new verbs, Persian requires to rely almost entirely on compound structures, rather than derivational processes. This explains why the majority of new verbal concepts are built in the form of complex predicates consisting of a light verb (LV) plus a pre-verb component (e.g. nouns, adjectives, or prepositional phrases) which carries the main semantic load of a new light verb construction (LVC). Given the propositions as such, this paper will examine a group of LVCs that are formed by the Persian LV dādan ‘to give’ as in a NP+LV combination. The main objective, thus, is to demonstrate how the semantic structure of heavy dādan is reflected in LVCs. The paper also aims to uncover the underlying motives behind the construction of such LVCs.

Section 2 introduces LVs and provides a brief description of the literature. In section 3, different types of causatives in Persian are discussed. Data and methods are presented in section 4. Section 5 deals with analysis of the data in relation to the meaning of dādan and section 6 discusses the findings and concludes the article.

2. Theoretical preliminaries

LVs have been the subject of a large portion of research since the term was coined by Jespersen (1940). He argues that LVCs are a way to add descriptive information in the form of an adjunct, as in “take a hot shower”. This means that LVs provide little or no meaning for the sentence, but it is the pre-verb component that carries the main semantic load.

Early studies approached LVs mainly through a syntactic perspective (for instance, see Cattell, 1984; Grimshaw & Mester, 1988; Mohammad and Karimi, 1992). In these studies, on the one hand, LV is considered as responsible for determining the argument structure of the sentence and, on the other hand, θ-roles are assigned to the arguments. By studying the Japanese LV, Grimshaw and Mester (1988) argue that the argument structure of suru ‘to do’ is incomplete; therefore, the pre-verb must lend or transfer its arguments to suru so as to make it a θ-marker. This line of argument was later adopted by Mohammad and Karimi (1992) in their study of Persian LVCs. However, the argument-transfer hypothesis fails when a pre-verb lacks any argument in the first place, such as the Persian noun āks ‘photo/picture’ in aks gereftan ‘take a picture’, or rāh ‘way’ in rāh dādan ‘allow to enter’. Considering both cases, Karimi (1997), then, concludes that both LVs and pre-verbs have meaning, and the majority of Persian LVCs are compositional in meaning.

In contrast to these syntactic accounts, cognitive linguists, by assigning a greater role to LVs, assume that there is a complex semantic relationship between LVs and their heavy counterparts. For example, Norvig and Lakoff (1987) provide a lexical network of different senses of heavy and light “take”. They argue that the associations among these senses are systematic and motivated by metonymy, metaphor, image-schema transformation, and profile shift, among other mechanisms. For example, the metaphor perceiving is receiving is active in “John took a whiff of the coffee”. According to these authors, differences in the various senses of “take” are minimal, and, at the same time, semantic. Newman (1996) also provides a comprehensive account of “giving” in English and other languages. He states that there are systematic relationships between various usages of heavy and light “give” that can be attributed to the complex semantic structure of the prototypical “give”. This structure, forming the basic action of giving (i.e. the giver’s transferring of an entity to a recipient), can be metaphorically extended to the situations where “give” expresses concepts such as permission, enablement, and causation, among others. Therefore, the metaphorical transfer of “control” maps to more abstract situations like, permitting as in the Persian LVC ejāze dādan [permission give] ‘allow’. This line of argument is adopted in this paper to show how LVCs are related to the semantic structure of dādan.

The fact that LVs’ contribution is more than syntactic features has received much attention in the literature. For example, Brugman (2001) argues that English LVs keep force-dynamic properties (in Talmy’s term, 2000) of their heavy counterparts, including lexical aspect (Aktionsart), and the semantic roles which are typically extended from physical to psychological domains. This means that if a heavy verb is (a)telic, its LV usage will also be (a)telic. This appears to be true for some LVs in other languages such as Persian, for example, gereftan ‘take’ and xordān ‘collide’ or ‘eat’, which primarily express telic, self-oriented events (Soltani et al., 2017; Soltani, 2018), but it cannot be generalized to other LVs (see Folli et al., 2005; Samvelian & Faghiri, 2014; Soltani, 2018). Moreover, LVs tend to maintain semantic roles in LVCs, providing further evidence that such verbs contribute to LVCs’ semantics (Newman, 1996; Brugman, 2001; Soltani, 2018).

Other researchers have also shown that LVs have meaning; for instance, Family (2006; 2008) examines the semantic space of Persian LVCs in the form of what she calls islands or “clusters of LVCs, expressing similar verbal notions, based on the same LV and a specific type of PV [pre-verb]” (Family, 2006: 50). She then argues that Persian LVs are not void of meaning. Following Goldberg (1995), Family continues that the semantics of such constructions is a function of the meaning of LV, pre-verb, and the construction itself; consequently, LVCs are not purely compositional in meaning. In this connection, Samvelian and Faghiri (2014) further expand the idea of LVC compositionality. They state that the idiomatic feature of many LVCs does not preclude them to be categorized based on their syntactic and semantic similarities - (e.g. sedā zadan ‘call’, dād zadan ‘shout’ and sut zadan ‘whistle’ - and that they are LVCs formed with zadan ‘hit’ and all refer to a general EMIT construction that describe deliberate production of various sounds), thereby resulting in compositional constructions. The important point, they argue, is that compositionality should be viewed a posteriori rather than a priori, implying that the meaning of an LVC cannot be obtained simply from a single component, such as LV or pre-verb.

This overview of LVCs shows that although syntactically grounded studies paid only scant attention to the semantics of LVs, the studies drawing upon cognitive frameworks have recognized crucial aspects of LVs meaning
 (Newman, 1996; Brugman, 2001; Family, 2006; Samvelian and Faghiri, 2014). The contribution of LVs varies from argument structure and semantic roles to the telicity of event structure (which is language- and even verb-specific), and even the general meaning of an LVC. Therefore, replacing one LV with another can completely change the meaning of an entire LVC, as in yād dādan [memory give] ‘teach’ versus yād gereftan [memory take] ‘learn’ or kotak zadan [beating hit] ‘beat someone’ versus kotak xordan ‘get beaten’. However, claiming that syntactic studies have entirely refused to account for LV’s meaning does not do justice to the contribution of this group of studies. For example, Karimi-Doostan (2001) states that LVs are responsible for aspectual properties or event roles in N+V complex predicates. He also suggests that LVs assign nominative and accusative cases (2005). He further continues by arguing that the (in)separability of LVCs is a function of the semantic and morphosyntactic properties of both the LV and the pre-verb (2011). Likewise, Megerdoomian (2001) maintains that aspect is related to LVs, and, in the same vein, Folli et al. (2005) argue that agentivity of the subject, the eventiveness, and duration of LVCs are associated with LVs.

It is now a well-established idea that not only have LVs meaning (Brugman, 2001), but they also contribute to the semantics of LVCs as well. Moreover, since LVs are derived from simple heavy verbs (Butt, 2010), it is natural for them to be semantically related to their heavy counterparts and retain various aspects of their semantics in LVCs. This may result in a phenomenon through which the LV acquires a new “specialized” function during the semantic bleaching process and is used to convey a particular concept. For example, the Persian LV xordan ‘collide/eat’[1] is mainly used in LVCs that express reception or undergoing as shown by the examples below:

1)

1. in tišert       rāhat      otu          mi-xor-ad.

     this     t-shirt      easy        iron         IPFV-collide.PRS-3SG

     ‘This t-shirt is easy to iron.’

1. havvā farib-e                    šeytān    rā            xord.

     Eve                   deception-EZ          Satan     OM           collide

     ‘Eve was deceived by the Devil.’

1. saqf-e xāne       tarak      xord.

     ceiling-EZ           house     crack      collide.PST.3SG

     ‘The ceiling of the house cracked.’

In (1.a), xordan in otu xordan is roughly the opposite of zadan in otu zadan ‘iron’ and has a reverse relation with it. Xordan conveys the concept reception as in kotak zadan / xordan ‘beat/get beaten’. This is used to tell whether clothes are easy or difficult to iron (otu naxordan[2] means ‘hard to iron’). Besides, the contact is physical in (1.a), whereas in (1.b) and (1.c) xordan no longer denotes such a contact but shows that the subject has received the result/outcome of someone else’s act (1.b), or a change in its state (1.c). These examples indicate that xordan often takes a “patient” or “undergoer” in the subject position in LVCs.

Similarly, the LV dādan functions as a causative maker in many of the resulting LVCs. This function is directly related to the semantic structure of heavy dādan such that many of the LVCs have a causative component to their meaning. The important point here is that the LV along with the pre-verb contributes to causation. In other words, the entire LVC must be considered as causative. This is consistent with previous studies that also highlight the role of construction, rather than a single element in LVC semantics (e.g. Goldberg, 1995; Family, 2006; Samvelian and Faghiri, 2014). Nonetheless, the semantic nature of the LV plays a significant role in the formation of causative LVCs, but this does not imply that all the resulting LVCs will be causative. By taking these preliminaries into consideration, we are now able to explore those types of LVCs formed by dādan and their underlying cognitive motivations on the one hand, and to argue for the causative status of the LV dādan in LVCs, on the other. As causation will be a major concept conveyed by dādan, we need to provide a discussion of certain important issues associated with it.

3. Causatives in Persian

To form causatives, languages use a variety of formal strategies ranging from morphological devices to periphrasis (Dixon, 2012: 242-49). In this connection, Persian causatives are briefly discussed here according to Dixon’s (2012) typology. The discussion includes morphological, lexical causatives, and syntactic (i.e. periphrasis) causatives:

3.1 Morphological causatives

As the label suggests, languages may use this type of causative by changing the morphological structure of the verb, including internal changes (e.g. in vowel quality), consonant repetition, vowel lengthening, reduplication, and affixation (Dixon, 2012: 242). Persian employs affixation by appending the suffix -ān to the present stem followed by the infinitive suffix -an. The suffix -ān transforms inchoatives (left) into causatives (right). Note that -d and -id turn the present stem into past stem to prepare it for the infinitive suffix -an:

(A)

1. xor-d-an (eat) >            xor-ān-d-an (feed)
2. xāb-id-an (sleep)                 >            xāb-ān-d-an (put to sleep)
3. xand-id-an (laugh) >            xand-ān-d-an (make laugh)
4. tars-id-an (be scared) >            tars-ān-d-an (frighten)
5. juš-id-an (boil) >            juš-ān-d-an (bring to boil)

2)

1. bače xābid.

     kid     sleep.PST.3SG

     ‘The kid has slept.’

1. mādar bače       rā            xāb-ān-d.

     mom                 kid          OM           sleep-CAUS-3SG.PST

     ‘Mom put the kid to sleep.’

3.2 Lexical causatives

In addition to the derivational process, there are two groups of causatives that do not require the affix ān. These include verbs that act as both inchoatives and causatives (B), and verbs whose causative counterpart is an intrinsically causative verb (C):

(B)

1. šekastan (break)                 >            šekastan (break)
2. poxtan (cook) >            poxtan (cook)
3. rixtan (pour) >            rixtan (spill)
4. boridan (cut) >            boridan (cut)

3)

1. livān xord zamin     va            šekast.

     glass  collide.PST.3SG        ground   and         break.PST.3SG

     ‘The glass hit the ground and broke.’

1. sārā livān       rā            šekast[3].

     Sarah                                glass       OM           break.PST.3SG

     ‘Sarah broke the glass.’

(C)

1. šodan (become) >            kardan (make)
2. oftādan (fall) >            andāxtan (fell/drop)
3. raftan (go) >            bordan (take)
4. āmadan (come) >            āvardan (bring)
5. mordan (die) >            koštan (kill)

4)

1. omid                 besiār     xošhāl    šod.

     Omid                 very        happy    become.PST.3SG

     ‘Omid became very happy.’

1. sārā omid       rā            besiār     xošhāl    kard.

     Sara   Omid      OM           very        happy    make.PST.3SG

     ‘Sara made Omid very happy.’

3.3 Periphrasis

A general way to form causatives in Persian is to use the LVC bāes šodan ‘cause’ in a superordinate clause followed by a complement clause. This is similar to “make” constructions in English with the causative verb “make” in the superordinate clause, as in “he makes her feel happy”:

5)

1. omid xošhāl    bud.

    Omid  happy    be.PST.3SG

    ‘Omid was happy.’

1. sārā bāes        šod                          omid       xošhāl    šav-ad.

     Sara   causer    become.PST.3SG      Omid      happy    SBJV.become.PRS-3SG

     ‘Sara made Omid very happy.’

In addition to these three types of causatives, there are both simple and compound verbs whose causative counterpart is a compound verb or an LVC. This type of lexical causative is frequently used with the verb kardan, and as previously stated, the light verb dādan can also be found in a significant number of lexical compound causatives:

(D)

1. didan (see) >            nešān dādan ‘sign give’ (show)
2. barxāstan (stand up) >            boland kardan ‘long do’ (make someone stand up)
3. qosse xordan ‘sorrow eat’ (sorrow)                 >            qosse dādan ‘sorrow give’ (grieve someone)
4. ātaš gereftan ‘fire get’ (catch fire)                 >            taš zadan ‘fire hit’ (set on fire)
5. sorat gereftan ‘speed get’ (speed up) >            sorat dādan ‘speed give’ (accelerate)
6. gul xordan ‘deceit collide’ (be fooled) >            gul zadan ‘deceit hit’ (deceive)
7. pāyān yāftan ‘end find’ (end) >            pāyān dādan ‘end give’ (put an end to)
8. anjām šodan ‘fulfilment become’ (get done) > njām dādan ‘fulfilment give’ (do)

6)

1. mājarā pāyān     yāft.

     controversy      end         find.PST.3SG

     ‘The controversy ended.’

1. ānhā mājārā                 rā            pāyān     dād-and.

     they   controversy           OM           end         give.PST-3PL

     ‘They put an end to the controversy.’

Such examples of LVCs show that the light verb dādan is the causative counterpart of LVs such as xordan ‘collide/eat’, gereftan ‘take’, yāftan ‘find’, and šodan ‘become’. This, however, does not imply that every LVC formed with these LVs can be turned into a causative using dādan, but rather that dādan can be a causative LV in Persian.

4. Methods and materials

In this study, more than 220 LVCs were collected from the two well-known Persian dictionaries, namely, Farhang-e Bozorg-e Sokhan (Sokhan Comprehensive Dictionary) (Anvari, 2003) and Farhang-e Farsi-e Emrooz (Contemporary Persian Dictionary) (Sadri Afshar et al., 1998). The data were obtained from the internet through browsing various inflected forms of dādan. The dataset also involves some instances derived from previous studies. It is worth noting that numerous unexamined novel LVCs can be found through browsing the internet webpages. Such new cases are usually coined by analogy with existing LVCs and may contribute to further lexicalization or grammaticalization of a concept.

A major challenge in dealing with LVCs is to distinguish them from other verb phrases such as direct object-verb combinations. As noted in the literature, dādan is one of the most frequent Persian verbs (Karimi-Doostan, 1997: 91-92), with a large number of LVCs[4]. To differentiate LVCs from non-LVCs, the data are restricted to LVCs formed with a noun as pre-verb (N+LV). Moreover, in most cases, LVCs have been selected according to the criteria proposed by Tabatabaei (2005) and Seifollahi and Tabibzadeh (2013) for differentiating complex predicates (CPs) from verbal phrases:

1- If the NP can be modified with quantifiers (e.g. čandtā ‘several’) and numerals (dotā ‘two’, yek ‘one’), the phrase is less likely to be an LVC:[5]

7)

1. dāvar                 be           u              yek          emtiāz    dād.

     referee               to            him/her  one         score       give.PST.3SG

     ‘The referee gave him/her one point.’

1. hame-ye mā           rā            farib       dād/                        *čandtā farib       dād.

     all-EZ                  we           OM           deceit     give.PST.3SG/           several   deceit     give.PST.3SG

     ‘S/he deceived all of us.’

2- If the direct object can be replaced with other nouns, the phrase is less likely to be an LVC:

8)

hedie / pādāš / jāyze / kādo dādan

to give gift / bonus / reward / present

3- If the unmarked (generic) object can be turned into direct object using the object marker rā, the phrase is less likely to be an LVC:[6]

9)

1. peyqām rā            be-de                      va            bi-ā.

     message            OM           SBJV-give.PRS.2SG    and         SBJV-come.PRS.2SG

     ‘Give the message and come back.’

1. xeyli left-eš                     mi-d-i / (*xeyli left rā midi)

    very    procrastination-it IPFV-give.PRS-2SG

    ‘You procrastinate too much.’

4- If the combination of NP and transitive verb requires a direct object, the phrase is more likely to be an LVC:

10)

qeymat-hā             rā            kāheš                      dād-and.

price-PL                   OM           decrease                                give.PST-3PL

‘They lowered the prices.’

5- If the combination of NP and transitive verb is an intransitive verb, the phrase is an LVC:[7]

11)

omid       dar          dādgāh                  šahādat                 dād.

Omid      in             court                       testimony              give.PST.3SG

‘Omid testified before the court.’

Although these criteria cannot draw a sharp distinction between complex and simplex predicates, they seem to show which word chain exhibits more CP-like characteristics. More precisely, apart from definite cases of LVC, namely N+V combinations with non-compositional or idiomatic meaning (in the sense of Nunberg et al., 1994 and Karimi, 1997)[8], borderline cases are also selected according to the above criteria. Moreover, linguistic intuition is also consulted because these criteria are not without flaws, as pointed out in footnotes (2) and (3). The criteria (4) and (5), however, seem to provide more reliable results.

The selected LVCs are described according to the principles of cognitive lexical semantics, including conceptual metaphor (Lakoff and Johnson, 1980; Lakoff and Turner, 1989), image-schema (Johnson, 1987, 1991, 1993; Lakoff, 1990) and foregrounding (Lakoff, 1987). Talmy’s (2000) notion of force-dynamics has also been used to show the force interactions in order to account for causative constructions formed with dādan. The use of dādan in LVCs conveying concepts ranging from physical to psychological causation can be explained by mappings between elements of physical transfer and those of a causative event, including possession, control, imposition, and reception among others. The image-schemas also help to explain how the elements of a prototypical event of giving can be the basis for forming new senses while showing minimal differences with related senses. The general idea behind these principles is that various senses of a word form a radial category (Lakoff, 1987) in which different senses are systematically linked to a central meaning and exhibit a prototype effect (Rosch and Mervis, 1975; Rosch, 1978). These links can be explained by the principles of cognitive lexical semantics. We then aim to show how the verb dādan is bleached semantically and used as a light verb, as well as how it is specialized in conveying a more nuanced notion, in this case causation.

5. Analysis of data

As a heavy verb, dādan is a ditransitive verb that requires a subject, an object, and an indirect object in a prototypical act of giving. Such a prototypical use depicts a kind of event where the subject or agent (NP1) takes an object (NP2), initiates an action, and transfers that object by hand to a recipient or an indirect object (NP3). This is shown in the argument structure of dādan:

NP1_(SUB) NP2_(DO) NP3_(IO) dādan

12)

omid       ketāb      rā            be            man        dād.

Omid      book       OM           to            me          give.PST.3SG

‘Omid gave me the book.’

This structure shows that the most fundamental concept articulated by dādan is the giver’s transferring of an object to a recipient. This can also be illustrated by using an image-schema (Figure 1) where the square represents the grammatical subject or giver (source), the dashed line rectangle is the direct object (theme) being transferred, the circle denotes an indirect object or a recipient (target), and the arrows indicate the direction of the object’s movement from the source to the target.[9]

Figure 1- The Image-schema of prototypical dādan

This schema is composed of three parts: (a) shows when the giver has the object in his control; (b) depicts transferring of the object; and (c) is the end-state of the event in which the recipient is in control of the object. As for the large rectangles, they show the domain over which one has control and is called sphere of control (Newman, 1996: 43). When giving something to another person, along with the transfer of an object, the giver may also transfer control over that object to the recipient, as in ejāze dādan [permission give] ‘permit’. It could be argued that the shift of control is part of the meaning of dādan and can be interpreted as a shift in the object’s position in relation to the sphere of control of elements involved in an event (Newman, 1996: 47).

Another dimension of the concept ‘transfer’ can be explained in terms of force-dynamics, which deals with notions such as energy and force interactions (Talmy, 2000; Brugman, 2001). As can be noted, the prototypical event of dādan requires at least three elements: a giver, an object or a transferee, and recipient. The basic force-dynamics of dādan can be described as follows: the giver is at the outset of energy flow, thus called the energy source, and the recipient is at the end of the energy flow, making it sink (Brugman, 2001). As such, the former is the initiator and agent of the action, while the latter is the beneficiary of the action’s outcome. The very agentivity of the subject and the act of transferring the object to the recipient are two fundamental elements that serve as the basis for metaphoric extensions and causative LVCs. This semantic structure could be entirely present, or parts of it could be foregrounded, with the rest being backgrounded. Therefore, the semantics of LVCs with dādan can be a function of all these components or a particular dimension of semantic structure of dādan. The following section discusses dādan LVCs and their relation to the semantic structure and force-dynamics of the heavy verb dādan. The subsections (a) to (m) are the semantic categories of the various LVCs examined in this article.

• Payments, financial transactions, (moral) compensation

The first group of LVCs are constructions in which dādan is closely related to the prototypical meaning of the heavy verb. In these LVCs, the pre-verb may be either the theme that is being transferred or the type of payment being made (e.g. rešve ‘bribe’; anām ‘tip’; bāj ‘ransom’) or a transfer being made (e.g. ejāre ‘rent’; rahn ‘mortgage’; vām ‘loan’):

(E)

Considering the semantic structure of dādan, it is interesting to note that force-dynamics is preserved in these LVCs in such a way that the transferee leaves the giver’s sphere of control and enters that of the recipient. This means that the transferee is literally given to a recipient. However, this is not always the case, as in kaffāre dādan ‘expiate’, ānd tāvān dādan ‘atone’; these verbs originally implied that someone is given things like money as compensation or in retaliation. They are also used figuratively to mean atone for one’s actions. Therefore, there is no third element as a recipient in such LVCs. This is also true for sadaqe / zakāt / xoms dādan which can also be used generically without any known recipients:

13)

tāvān-e                  gonāh-ān-am        rā            mi-dah-am.

atonement-EZ        sin-PL-1SG.POSS        OM           IPFV-give.PRS.1SG

‘I expiate my sins.’

14)

bāyad     zakāt-e   māl-at                    rā            be-dah-i.

must       zakat-EZ asset-2SG.POSS        OM           SBJV-give.PRS.2SG

‘You should pay the poor-rate of your assets.’

• Acts of communication, content and knowledge transfer

The second type of LVCs are similar to the LVCs in group (a) in that they both involve some sort of transfer between a giver and a recipient. They, however, differ in that the pre-verbs here designate various forms of communication, and therefore the transfer is more of a metaphorical kind, with the pre-verbs ranging from more concrete (alāmat ‘sign’) to more abstract (yād ‘memory’) notions where the noun could hardly be conceived as a transferee:

(F)

These LVCs refer to different types of interpersonal communication which require at least three elements to be present: an addresser, a content or a message, and an addressee[10]. Following Newman (1996: 138), the use of dādan in such LVCs can be explained by drawing on the conduit metaphor (Reddy, 1979; Lakoff and Johnson, 1980: 9-12). Based on this type of metaphor which is made up of three components, namely ideas are objects, linguistic expressions are containers, and communication is sending (Lakoff and Johnson, 1980: 9), the addresser puts ideas in the words, and transmits them through the language to the addressee, who then decodes and understands them. Similarly, these elements could be described by mappings of the addresser onto the giver, the communicative content onto the transferee, and the addressee onto the recipient (Figure 2).

Figure 2- Metaphorical mappings between prototypical dādan and interpersonal communication

Based on such mappings, možde ‘tidings’, salām ‘hello’, tazakor ‘admonition’, šahādat ‘testimony’, and similar forms that refer to communicative content are conceived as transferrable objects. Even in a more recent LVC, like konferāns dādan ‘lecture’ (possibly in analogy with erāe dādan ‘present’, or dars dādan ‘teach’, or a loan translation from English (i.e. give a conference), the transfer of communicative content in the form of speech is evident. A further point relates to verbs like yād dādan ‘teach’ or erāe dādan ‘present’, where the pre-verbs do not represent content (unlike āmuzeš ‘teaching’), but the entire LVCs require a second object referring to content. This is not to say that other LVCs do not require a direct object, but that it implies that in a verb like yād dādan, where dādan is probably lighter than in āmuzeš dādan, we can use the pre-verb as a direct object and generate sentence (15) but not for yād dādan or erāe dādan (16).

15)

āmuzeš-hā-i          ke            man        mi-dah-am.

teaching-PL-REL     that         I              IPFV-give.PRS.1SG

‘Things that I teach.

16)

*yād-hā-i              ke            man        mi-dah-am.

memory-PL-REL     that         I              IPFV-give.PRS.1SG

The ability of some pre-verbs to become direct objects suggests that associated LVCs are most possibly formed through noun incorporation (Dabir-Moghaddam, 1997); it can then be used as a criterion to determine the degree of lightness of a given LV. This, however, requires further investigation of various LVCs with different LVs, which is beyond the scope of the present paper.

• Causation – state of body and mind – positive

The verbs in this category represent various types of physical and psychological events in the sense that a given LVC can be used with a subject that could be either an action (17), a state or situation (18) or an entity, but the result is often a condition of the body that ultimately creates a particular mental state in the recipient:

(G)

17)

davidan                 be            man        ārāmeš                   mi-dah-ad.

running                  to            I              tranquility             IPFV-give.PRS-3SG

‘Running comforts me.’

18)

čašm-ān-e              u              be            man        ārāmeš                   mi-dah-ad.

eye-PL-EZ                                s/he         to            I              tranquility             IPFV-give.PRS.3SG

‘His/her eyes comfort me.’

As previously stated, a major element of giving is transfer. Note that this transfer is not physical here, but there is a mapping from the concrete onto the abstract world. To put it differently, the transferee is an emotion or mental state that the sender transfers to the recipient. This mapping is captured by the metaphor states are objects (Kövecses, 2008), which maintains that due to an action or situation (the energy source), abstract notions like the states of mind or feelings are conceived in terms of transferable objects moving from the giver (the energy source) to the recipient (the energy sink). This mapping is depicted in Figure (3):

Figure 3- Transfer of states or emotions

What makes LV different from heavy dādan in these LVCs is that, with the literal meaning, the sender has the transferee in his sphere of control (part (a) of Figure 1) and then gives it away, while, as Figure (3) shows, the sender does not possess the transferee, but induces a feeling he may or may not have himself. In other words, the giver is the source of a feeling; this source may be a condition (sleeping), an event (a party), or an action performed by a person (talking). As a result, the recipient or energy sink will be in a state of mind s/he did not expect before.

• Causation – state of body and mind – negative

The reason these verbs are in a different category from the verbs in (c) despite the same underlying process (emotion/mental state as the pre-verbs), is the difference in the type of concepts they convey. The LVCs in group (c) express mainly positive emotions, whereas the following LVCs cause the recipient to suffer.

(H)

Here, the giver or initiator of the action causes the recipient to experience a condition or situation. In other words, the recipient’s experience (e.g. torment) is the result of the giver’s action, which is transferred to him/her. This is even more evident in gorosnegi or tešnegi dādan where the agent causes the recipient to suffer by actually “not giving”.

• Causation – accomplishment – status change

Another group of LVCs with causation at their core consist of verbs that convey more tangible or physical concepts, as opposed to the previous causatives that often denote abstract and psychological notions:

(I)

The pre-verbs here denote the state of affairs of the recipient, resulting from an action on behalf of the subject. In other words, the subject does something whose effect is transferred and imposed on the other party in the form of these pre-verbs. The source of energy can be either a causing agent (19) or a causing event (20) as in other causative LVCs:

19)

modir-e                  jadid                      be            kār-hā                    šetāb                      dād.

boss-EZ                   new                         to            work-PL                  acceleration          give.PST.3SG

‘The new boss speeded things up.’

20)

tavarrom               qeymat-hā             rā            afzāyeš                   mi-dah-ad.

inflation                 price-PL                   OM           increase                 IPFV-give.PRS-3SG

‘Inflation will raise prices.’

Virāž dādan (for cars), mānovr dādan (for operations) and jolān dādan (for horses) are now mostly used intransitively through metonymical extension (rider for a moving vehicle), with the latter having a new figurative meaning (to govern or rule). As causative LVCs, they could be conceived as giving the status of movement to the recipient like in jolān dādan where jolān meant the status of a running animal (usually a horse).

• Causation – physical change

The following LVCs also denote change, but the change is physical and visible. The pre-verbs mainly refer to physical attributes that describe the kind of action the subject or agent does to the recipient (mostly inanimate):

(J)

Nevertheless, as in the following sentences, some of these verbs can convey more abstract and figurative changes in an animate recipient, wherein jalā dādan means both make something gleaming in look (21) and purify one’s soul (22). Even in these cases, causation is evident:

21)

hame-ye                 goldān-hā             rā            jalā         dād-am.

all-EZ                       vase-PL                   OM           gleam     give.PST-1SG

‘I made all the vases glossy.’

22)

adabiāt                  ruh-e      ensān     rā            jalā         mi-dah-ad.

literature                soul-EZ    human   OM           gleam     IPFV-give.PRS-3SG

‘Literature purifies human soul.’

• Causation – motion (literal and figurative)

Motion verbs are the most salient types of verbs (Miller and Johnson-Laird, 1976: 527) and constitute a significant part of the human lexicon (Diller, 1991). The fact that dādan has ‘transfer’ as part of its meaning makes it an appropriate candidate for forming LVCs with motion-related pre-verbs:

(K)

The LVCs are examples of caused-motion constructions (Goldberg, 1995) in which the subject causes the recipient of the force to move, and the pre-verbs specify the type of movement. In other words, the subject is the agent of a caused-motion event (Kodama, 2004). Motion can range from directly moving the recipient (harekat dādan) to a less direct movement, but more apparent causation (rahāi/kuč dādan). Figurative movement is still implied when the caused movement is less evident, as in rahāi dādan (24). This is mainly because freeing means ‘causing or helping someone to move out of an enclosed area’ (whether physical or psychological confinement):

22)

bāyad     say          kon-i                       pā-yat                    rā            harekat                  be-dah-i.

must       try           SBJV.do.PRS.2SG       foot-2SG.POSS          OM           movement            SBJV-give.PRS-2SG

‘You should try to move your foot.’

24)

u              rā            az            miān-e                    afkār-aš                 rahāi      dād.

s/he         OM           from       between-EZ            thoughts-3SG.POSS  freedom give.PST.3SG

‘It freed/unleashed him/her of his/her thoughts.’

• Causation – expansion – spatial and temporal

The following LVCs are additional examples of causative constructions sharing similarities with some of the previous LVCs discussed so far. For example, they exhibit both accomplishment-like properties like sorat dādan ‘speed up’, and motion-like characteristics analogous to those of the previous category in (g). The difference, however, lies in the pre-verbs that denote different types of spatio-temporal expansion:

(L)

The agent causes the recipient on the energy path (the direct object) to expand either spatially (e.g. vosat dādan ‘expand’) or temporally (e.g. edāme dādan ‘continue’). The expansion is conceptualized in terms of transferring and giving the possession to the recipient, much like the metaphor states are objects. However, this does not imply possession on the part of the subject or agent. Rather, it suggests that the agent causes the recipient to be in a state where it has not been before or transfers and imposes the state onto the recipient.

Figure 4- Causation and expansion of sphere of control

Additionally, expansion entails movement from a source location onward, thus these LVCs can be viewed as examples of figurative motion in which a single entity (e.g. a company, theory, etc.) is conceptualized to continue moving onward and cover larger areas (e.g. the expanded sphere of control in Figure (4) depicts caused expansion). Moreover, movement in space implies movement in time and, therefore, expansion can be used figuratively for prolongation of time:

25)

efrāti-un                                dāyere-ye              amal-e                    xod         rā            vosat      dād-and.

extremist-PL           circle-EZ                  function-EZ            self          OM           extent     give.PST-3PL

‘The extremists expanded their circle of activity.’

26)

talāš-am                                rā            edāme                    mi-dah-am.

effort-1SG.POSS       OM           continuance          IPFV-give.PRS-1SG

‘I keep up my efforts.’

• Causation – linking

The following LVCs are instances of relations, links, connections, or bonds between entities conceptualized as a state-of-being linked, given (caused) by the sender to the recipient. However, what distinguishes these LVCs is the requirement of at least two entities for the link or connection to happen. Figure (5) depicts this type of causation.

(M)

Figure 5- Causative linking

According to Figure (5), each recipient has his/her sphere of control and by, metaphorically giving the transferee, the sender or agent causes him/her to share the same sphere of control over the transferee, thereby creating a connection between the two (27). Same as the group (h), the agent is not a possessor, but causes others to be in a particular state. Although the image-schema holds when both the recipients are human (or animate at least), as they can have control over their affairs, it can be extended to other scenarios where the recipients at the other side of the force path are not human (28a) or even animate (28b).

27)

zan          va            šohar      rā            āšti                         dād-and.

wife        and         husband                OM           reconciliation        give.PST-3PL

‘They made the couple reconcile.’

28)

1. u rā            bā           adabiāt                  ons                          dād-am.

       s/he  OM           with        literature                fondness                                give.PST-1SG

       ‘I made him/her fond of literature.’

1. zamin o              zamān                    ro            be            ham                        rabt        na-deh.

       earth                and         time                        OM           to            each other             relation  NEG-give.PRS.2SG

       ‘Don’t try to relate everything [unrelated things].’

• Enablement – permission – shared control

To have an object in one’s sphere of control implies the exertion of control over it. Besides, by transferring an object to the recipient, the sender also may transfer the control over that object. Moreover, when someone is in control of something, s/he may be able to do things that could not be done otherwise (Newman, 1996: 182). In other words, by giving control to the recipient, the sender permits or enables him/her to act upon it. The following LVCs are examples of events in which the agent enables or allows an entity (mainly human) to perform an activity:

(N)

LVCs such as ejāze / roxsat / forsat dādan, in which the pre-verbs virtually mean permission, indicate the semantic compatibility of dādan with this concept. Furthermore, if one has meydān ‘arena’, rāh ‘way’ or bāzi ‘game’ in his possession[15], he has control over it. Importantly, the control is implemented through metaphorical transfer. But as the pre-verbs imply, this transfer is more of a sharing, than simply relinquishing one’s control, because one can deprive the recipient of such privileges at his/her own will (e.g. šerkat dādan ‘include, make involved’ is a prime example). Figure (6) illustrates a shared sphere of control similar to Figure (5) but, instead of two recipients, it is between the sender and the recipient. Ehtemāl dādan is an exception to Figure (5), because probability is not in anyone’s control. Therefore, although the giver has no control over it, he attributes it to a phenomenon (roughly corresponds to the recipient of the force) based on his assumptions to supposedly make it happen.

Figure 6- Sharing the control in permission and enablement

29)

dolat                      be            javān-hā                                meydān                  ne-mi-dah-ad.

government          to            young-PL                                arena                      NEG-IPFV-give.PRS-3SG

‘The government does not give a chance to the youth.’

30)

lotfan     u              rā            ham        bāzi        be-dah-id.

please     s/he         OM           too          game      SBJV-give.PRS-2PL

‘Let him/her play with you, please.’

A question that may arise is how to differentiate causation from enablement, the answer to which is not particularly clear for Persian. Copley et al. (2015) argue that in causation, the agent’s force is greater and not in concordance with the patient’s force, while in enablement, both the agent and patient exert force toward the same direction or end-state. However, Persian does not seem to encode the distinction, and context is needed to determine if a verb like sor’at dādan ‘increase’ is causing or enabling an event. However, enablement can be seen as a subtype of causation in that the sender causes the transfer of control to the recipient.

• Emission – exposure

The sphere of control also accounts for another group of LVCs. As was pointed at the beginning of this section, the transferee leaves the giver’s sphere of control and enters that of the recipient (Newman, 1996: 144). In the LVCs below, the exit from the sphere of control is mapped metaphorically on the exit from a container, resulting in exposure to a third party. Therefore, coming out and being exposed seem to be the basis for the following LVCs:

(O)

(P)

The first group are events where something comes out of a source location[18] (e.g. bu ‘smell’ of food; mive ‘fruit’ of tree). Therefore, they mean emit or produce a material. The LVC birun dādan differs from others in that birun does not denote a transferee but coming out of the transferee from its earlier location (the difference is in transitivity). Others are more figurative as in samar/natije dādan where giving results relates to giving fruits, or šekam dādan in which a big stomach is the basis for being saggy (as in ceilings or wall). Producing or making something is also active in suti / gāf dādan, both meaning gaffe (to make a gaffe), while the element of exposure (becoming embarrassed in front of others) is evident as well.

The second group LVCs are also close in meaning to emission, but the focus is more on exposure. Two interesting examples are rox/ruy dādan [face give], which idiomatically mean “to happen”. When a phenomenon happens, the phenomenon or its impacts will become exposed and visible to others, which is probably the basis for such constructions.[19] Figure (7) gives a graphical view of the above LVCs. Unlike the previous verbs that often require a recipient, the emphasis here is on the transferee’s leaving the sphere of control of the source location, and therefore it does not necessarily end up in the recipient’s sphere of control. Nevertheless, a recipient is understood to exist in the background as the exposure could lead to entrance to the perceptual or cognitive field available to the recipient.

Figure 7- Emission – exposure

Emission is mostly involuntary[20] (31) because the subject cannot exert force in the sense that humans do (even in suti/gāf dādan, the human subject cannot control their action), but exposure can happen both intentionally (namāyeš dādan) (32) and unintentionally (rox dādan).

31)

havās-aš                                na-bud                   ke            suti         dād-e.

attention-3SG.POSS NEG-be.PST.3SG        that         gaffe      give.PST-PRF

‘S/he didn’t notice he/she had made a gaffe.’

32)

honar-e                  xod         rā            be            hame      nešān     dād.

art-EZ                      self          OM           to            all            sign         give.PST.3SG

‘S/he showed his/her art/skills to everyone.’

• Give up control – submission

The following LVCs have body parts or associated notions (e.g. savāri > to lift and carry/ride) as their pre-verbs. Everyone is in control of their body, and giving up body or a body part, although metaphorically, means losing the control over it. Therefore, the metaphor control is possession (Gibbs et al., 1997) is primarily at work, much like other LVCs. That is why these LVCs are also closely related to the concept permission:

(Q)

33)

doxtar-e                 belaxare                                be-š         pā           dād.

girl-DEF                    finally                    to-him    foot        give.PST.3SG

‘The girl finally gave up [and agreed to go out with him].’

• Addition - implementation

The following LVCs denote two types of actions but appear to be on a continuum. At one extreme, the verbs are instances of adding the transferee (the pre-verb) to a recipient. The transferee is a material (dud), a form of energy (harārat, fešār), or even an end-state (juš). At the other extreme, the transferee is the activity or the end-state (varzeš, tamrin) performed as the result of the subject’s action:

(R)

There are also middle cases that exhibit attributes of both ends; for instance, māleš/māsāž dādan ‘rub, massage’ are close to fešār dādan ‘press/push’ and narmeš dādan ‘flex’ in that they both encode addition of a force and, at the same time, the end-state of the implementation of the force. In all these LVCs, however, the object which is at the end of the energy path is the beneficiary of the action. Therefore, some of these verbs also behave as causatives in certain contexts, where the object is not merely passive and can act on the force transferred from the subject (35):

34)

dast-am                  rā            māsāž                     dād-am.

hand-1SG.POSS        OM           massage                                give.PST-1SG

‘I massaged/rubbed my hand.’

35)

hāmeye                  bače-hā rā            tamrin    dād-am.

all-EZ                       kid-PL      OM           train        give.PST-1SG

‘I exercised all the children.’

The list of LVCs discussed so far is not exhaustive, but it reflects the processes underpinning the formation of different concepts using dādan. However, other LVCs may not fall into the above categories, and yet share some of their properties. One example is tahvil [delivery] dādan ‘deliver’ which is very similar to prototypical dādan. Others, like the list below, are less obvious but show that the subject’s action is beneficial to the recipient. In other words, the recipient receives the benefits from the giver’s action:

(S)

Newman (1996: 51-52) argues that the typical scenario of giving has a benefactive effect on the recipient, like the above examples, but, on the other hand, a large part of the LVCs with dādan act neutral or have a negative impact on the recipient, especially in causative constructions. LVCs like gušmāli [ear-rubbing] dādan ‘punish’, or gir [trap] dādan ‘pick on’, again not fitting the groupings, also show a negative impact imposed on the recipient. This might be attributed to the fact that causation can be both positive and negative.

6. Discussion and Conclusion

The analysis of the heavy verb dādan has shown that it has a straightforward semantic structure consisting of a giver, who is in possession (and thus in control) of an entity (transferee) and transfers it to a recipient, who then will be in possession and control of the transferee. Along with other cognitive processes, this structure, can account for a wide variety of LVCs, ranging from more literal (rešve dādan ‘bribe’) to highly figurative instances (vosat dādan ‘expand’). The data also show that dādan can be used as an LV in several ways. Firstly, the most literal LVCs roughly correspond to the image-schema of heavy dādan (Figure 1) in that they retain its semantic structure. For example, financial transactions (e.g. ejāre dādan) involve a physical give-and-take, and by semantic extension, a metaphorical transfer of possession and control (see the green line in Figure (9) connecting box (a) to control). The same schema applies to linguistic communications in which there are giver-addresser, transferee-content, and recipient-addressee mappings, though the physical transfer is not always present. This is also true of other languages like English, German, Italian, Bulgarian (Newman, 1996: 141-142), and French (Bouveret, 2012), suggesting that this is a cognitively motivated process.

Second, a great deal of the LVCs, as proposed in this paper, convey causation. Seven groups of causative LVCs were identified, each with its own set of idiosyncratic properties, including positive or negative impact, physical or mental change, motion, expansion, or linking. These causative LVCs can be accounted for as follows: the metaphor states are objects is present in all these groups, shown by the blue links between box (c) and other causative groups in Figure (9). States (e.g. speed, change, torture, expansion, etc.) are construed as transferable entities given to target recipients, making them have those states. Nolan (2015) also argues that transfer is the basis for using Irish thug ‘give’ in causative constructions, and Newman (1996: 173-175) has shown the same for Finnish, Thai, Polish, Cambodian, and Mandarin. Bouveret (2012) also states that French donner ‘give’ is used to communicate causative constructions. Another aspect is the force-dynamics of dādan; both causation and dādan depict other-oriented actions (Brugman, 2001), where the flow of energy moves from the giver to the recipient. Even in LVCs without a recipient, the energy flow moves away from the giver or the grammatical subject. Another important point about causative LVCs is that, as illustrated in Figure (3), causation does not require the subject to possess the transferee, but to be the source of the energy that causes the recipient to be in a particular state. Thus, the image-schema in Figure (3), with the giver’s possession eliminated, underpins the causative LVCs.

Figure 8- dādan and causation

Figure (8) summarizes how dādan is used in causative LVCs. It is worth noting that the overarching term causation does not imply that the causative LVCs are of the same type. The distinction between intentional and unintentional or agent and author causation, event or instrument causation, etc., as termed by Talmy (2000: 514-16), is beyond the scope of this paper. More fine-grained distinctions, however, would provide valuable insight into the nature of such LVCs.

The third aspect is the notion of control and possession (and the metaphor control is possession), which are specifically active in LVCs that denote: (a) expansion (as the sphere of control expands spatially and temporally), (b) linking where there are at least two recipients at the end of the energy path made to share a sphere of control, (c) permission, and enablement, in which both giver and sender would share control over an event, (d) submission, in which the subject would give up control and surrender to a third party, and finally (e) exposure, where coming out of the sphere of control is foregrounded. Therefore, the sphere of control can be expanded, shared, or given up on, resulting in LVCs that denote these and related concepts. Control is also relevant for causation, as was shown in the analysis. Fujiwara et al. (2014) also maintain that possession is the reason why “have” is used causatively in English, arguing that controllability of an event is required for “have” to be used as a causative verb.

The types of concepts represented by the LVCs are illustrated in a radial category in Figure (9). The LVCs are either directly related to dādan (showing close association with the heavy verb) or grouped around the two concepts causation and control. Box (c) interconnects, through the blue lines, to the other causatives showing the ubiquity of states are objects metaphor. The red lines indicate causative usage, and the green ones highlight the role of control and possession in LVC formation. The Figure shows that, despite being divided into different groups, the LVCs form various paths or continuums of interrelated concepts. For example, tārtib dādan ‘order’ and šekl dādan ‘form’ in (e) and (f), ’ozlat dādan ‘isolate’ and rahāi dādan ‘free’ in (e) and (g), or serāyat dādan ‘transmit/contaminate’ and entešār dādan ‘propagate’ in (g) and (h) respectively appear to be closely related but put into different groups. It is, however, beyond the scope of this paper to determine accurately which group an LVC belongs to or how the continuums interconnect because of the semantic nuances between even two closely related concepts. For example, although sor’at dādan ‘speed up’ is a motion verb, and thus a candidate for caused-motion LVCs (group g), the difference lies in the fact that it implies an earlier state of motion, while the LVCs in the group (g) do not assume earlier motion. The issue is also addressed by Collins (2015), who introduces “semantic distances” or intermediating related senses that come between two distinct senses of a word and bridge them across a continuum.

(A) Payments-financial transactions-(moral) compensation; (B) Acts of communication-content and knowledge transfer; (C) Causation-state of body and mind-positive; (D) Causation-state of body and mind-negative; (E) Causation-accomplishment-status change; (F) Causation-physical change; (G) Causation-motion (literal and figurative); (H) Causation-expansion-spatial and temporal; (I) Causation-linking; (J) Enablement-permission-shared control; (K) Emission-exposure; (L) Give up control-submission; (M) Addition-implementation.

Figure 9- Radial category of dādan LVCs

A final point regarding the analysis of the LVCs is the use of image-schemas. Since the LVCs are conceptually interrelated to each other and to the heavy verb, it is reasonable to expect that the image-schemas will be related to a core image-schema, which is exactly what happens in our analysis. All the image-schemas are related to the image-schema (1), and Figure (3) is the basis for the causative LVCs including the image-schemas (4) and (5). Figure (1) shows the path of dādan, while Figure (3) and causative LVCs focus on the end of this path. It is, therefore, safe to assume an end-point focus or end-of-path image-schema transformation (Lakoff, 1987: 423-4 and 441), focusing on the end-state of the recipient, while Figure (7) has a start-point focus highlighting the beginning of the giving process.

To summarize, the image-schemas are related to the main schema through the elimination of the source location or destination of the transferee, addition of at least two recipients, and sharing the transferee and the sphere of control. However, these findings do not imply that all the LVCs can be accounted for in this way. Analogy is another important factor in work in language change, especially when a given construction becomes frequent enough in language. In this way newer forms become generated based on frequent existing forms such that they cannot be explained in the same manner as those basis forms. Therefore, one should always keep in mind that not all LVCs are describable with a predetermined set of rules and principles.

In general, this paper shows that LVCs can be accounted for by the principles of cognitive lexical semantics, and that many of the dādan LVCs are causative, suggesting that it functions as a causative verb in Persian. In fact, dādan is sometimes used in situations where an LVC with kardan ‘do, perform’ is not causative as in rošd kardan/dādan ‘grow/make grow’, tamrin kardan/dādan ‘exercise/train’, obur kardan/dādan ‘cross/make cross’, etc. This process is still active and visible in rather new LVCs like xande [laughter]/gerye [weeping] dādan ‘make laugh/cry’, boqz [sob] dādan ‘make one’s voice look sad’ or the completely new and highly ironic LVC xodkoši [suicide] dādan ‘make someone suicide’ as opposed to xodkoši kardan ‘suicide’.

Abbreviations

[1] xordan is a homonymous word, it can mean either ‘collide’ or ‘eat’.

[2] na- negates the verb.

[3] Some inchoative verbs like šekastan and boridan can also take the causative suffix, an as in šekāndan and borāndan respectively.

[4] “Give” is also among the most frequent LVs in English (Tu and Roth, 2011) and it is used more as an LV than a heavy verb (Wittenberg et al., 2014).

[5] These criteria are not always decisive, and even definite cases of LVCs can override them as in hol dādan ‘push’ > do tā hol bede [two CLF push SBJV-give.PRS.2SG] ‘Give a couple of pushes’ or fekr kardan ‘think’ > az dišab hezār tā fekr kardam [from last night thousand thought do.PST.1SG] ‘A thousand thoughts have crossed my mind since last night’.

[6] It will fail when the pre-verb is modified as in fekr kardan ‘think’ > in fekr rā kardam ke … [this thought OM do.PST.1SG that] ‘I had a thinking that …’. Many of the pre-verbs in definite LVCs can be used this way.

[7] Four and five are mutually exclusive. However, some LVCs may not follow this like sor’at dādan ‘speed up’ > be kār-hā sor’at dād [to work-PL speed give.PST.3SG] / kār-hā rā sor’at dād [work-PL OM speed give.PST.3SG] ‘He speeded things up’. This may further support the bleaching of dādan in cases with a direct, instead of an indirect object.

[8] Karimi argues that many Persian LVCs are idiomatically combining expressions.

[9] NP labeling is added for clarification and clearly not part of the schema.

[10] And of course, the medium of communication which does not concern the present paper.

[11] This is an archaic usage which means to cause or help someone to be ahead of others. It is synonymous with sebqat dādan (also archaic).

[12] Informal use of qalt dādan; both mean to roll something.

[13] Other synonyms with the same structure are vaqt (time)/zamān (time)/mohlat (deadline) dādan meaning to give (more) time.

[14] The more accurate English equivalence would be to think of something as probable.

[15] Expressions like meydān-dār ‘arena-possessor’ and sāheb-extiyār ‘possessor-(of)-authority’, both meaning ‘to be in charge of’, suggest that control is possession.

[16] Samar originally means fruit, but it is now often used to mean result.

[17] From French pose (posture).

[18] This implies fictive motion, much the same as expansion.

[19] Or they might be truncated versions of rox/ruy nešān dādan.

[20] birun dādan can be voluntary for humans.

[21] A seemingly more recent LVC, very probably formed by analogy to kešik dādan (the word kešik originally referring to Mongolian royal guards).