---

Adding suffix is one of the methods of making words to produce various meanings. Various semantic contexts vary depending on the suffixes and prefixes in the word derivation. The suffixes "-är", "-gär", "-gar", "-an", "-nadeh" are known in Persian as suffixes that produce adjectives and nouns. In Russian the following suffixes:
-тель, -арь, -чик, -щик, ...
In the row, there are suffixes for making nouns. Among the above two suffixes "-арь" and "-тель" in Russian language are analyzed to find equivalence in Persian language.
The purpose of writing this research is to identify the active and passive suffixes in Russian and Persian languages ​​and to examine them against each other in understanding their meaning and form, relying on the theory of semantic-lexical expression (Liber) 2004 and the grammatical theory. The questions that we seek to answer in this article are, what is the degree of convergence and divergence of the semantic expansion of active and passive suffixes in two languages? Is it possible to make some words with the help of existing words, according to the generative and cognitive grammar approach, to what extent the semantic overlap allows the reproduction of some figures. And as a result of this research: it was found that the investigated suffixes in two languages ​​use the same basic units of noun + infinitive + suffix, but in the semantic part, the expected correspondence does not match completely.
 

---

Introduction
The dependents of verb are among the most debated subjects on which a considerable body of research has been done. Yet, researchers have constantly had diverse opinions about their real identities. Complement, as one of the dependents of verb, is in the same boat. Some scholars have differentiated obligatory complements from optional ones, while others consider complements as obligatory elements and do not recognize an optional category. This article, based on Langacker’s (1987, 2013) Cognitive Grammar and through a corpus-based method, seeks to find out whether the Persian corpus verifies the existence of optional complements and if not, in what category can we place what is normally called optional complement. In other words, this research is to seek the answers to the following questions: Are there any optional complements besides obligatory ones based on Persian corpus-based data as well as Langacker’s Cognitive Grammar? If complements are merely obligatory, how can one categorize those elements called optional complements?
Methodology
To answer the above-mentioned questions, four dependents (subject, object, source and goal) of four salient motion verbs (raftan 'go', āmadan 'come', āvardan 'bring' and bordan 'take') in Persian were chosen to be studied. To this end, 300 tokens of each salient motion verb along with their dependents and the related linguistic context were randomly selected from the corpus of Hamshahri 2 to observe their corporal behavior.
Discussion
Langacker (1987, 2013) distinguishes 3 dependents for heads including verbs, which are “complements”, “modifiers” and “adjuncts”. He defines complements as “a component structure that elaborates a salient substructure of the head. The head is thus dependent, and the complement is autonomous” (Langacker, 2013: 203). Conversely a modifier is “a component structure that contains a salient substructure elaborated by the head. In this case the head is autonomous, and the modifier is dependent” (Langacker, 2013: 203). And finally “a component structure which fails to either elaborate the head or be elaborated by it is called an adjunct” (Langacker, 2013: 205).
Regarding the four dependents of the salient motion verbs under study, subjects and objects are complements since they elaborate the salient substructures of the verbs. Subjects elaborate the schematic trajectors of the verbs and objects elaborate the schematic landmarks of them. So the verb is, to a great extent, dependent on the subject and the object to complete its meaning. Such high conceptual dependence of the verb brings about its syntactic dependence too and as a result complements are obligatory and must constantly accompany the verb. The corporal behavior of the complements (subjects and objects) verifies this fact; from 300 tokens of each verb in Persian, there was not even a single sample in which the subject or the object was absent. Goals and sources, which tend to be considered as optional complements in the canonical viewpoints in Persian grammar, are, taking Langacker’s Cognitive Grammar into consideration,  modifiers since the motion verb elaborates their schematic trajectors which is a schematic process denoting a motional action. As a result, they are conceptually dependent on the motion verbs, hence being modifiers.
3. Conclusion
The corporal behavior of subjects, objects, goals and sources as the dependents of the four salient motion verbs under study produces the following conclusions:

1. Complements are solely obligatory elements since they elaborate the schematic trajectors or landmarks of motion verbs; thus, motion verbs are so conceptually dependent on the complements that they can never appear without them and as a result they become syntactically dependent on the complements as well. Sources and goals, on the other hand, are modifiers that are dependent on motion verbs to elaborate their schematic trajectors. Therefore, the relation that exists between the complement and the verb does exist between the modifier and the verb too but in a reverse direction.
2. Although sources and goals are both modifiers considering Langacker’s Cognitive Grammar, the result of the study shows that there is a goal over source preference. The frequency of the goals is much higher than that of the sources and the result of the Chi-square test indicates that there is a significant difference between the presence of these two elements with salient motion verbs (P<0.05). This result aligns with Stefanowitsch and Rohde (2004), Kabata (2013) and Verkerk (2014).
3. Although there is an asymmetrical distribution between sources and goals, neither of them are optional elements. Their behavior in the text corpus shows that the presence of these modifiers are determined by the context, i.e. if the context needs them, they have to appear and if not, they are not employed by it. For that reason, sources and goals are contextually obligatory and can be called “contextual supplements”.

Studying adjuncts in the corpus shows that they are not optional either. These elements, too, have to be present if the context necessitates their being but if they are not summoned by the context, they are absent. So, adjuncts on the par with the modifiers are contextually obligatory and termed “contextual supplements” in this study. Based on the results of the analysis of the Persian text corpus, it seems that Langacker’s triple division of the dependents (i.e. complements, modifiers and adjuncts) does not meet the corporal behavior of these dependents.

---

Introduction
Affixation is one of the two main word-formation processes in Persian. The suffix ‘-ɑr’ is a nominal (or adjectival) suffix in Persian which is believed by most researchers to express the subject/agent (like xæridɑr ‘buyer’), object/patient (like gereftɑr ‘captive’) or infinitive (like ræftar ‘behavior’) meaning. In Cognitive Grammar (Langacker, 2008, 2009) and Construction Morphology (Booij, 2010, 2016) word-formation patterns are considered to be constructional schemas, i.e. schematic representations of morphological constructions. Constructions are pairings of form and meaning. The form pole of a construction includes morpho−syntactic and phonological properties. The meaning pole of a construction comprises semantic properties (conceptual structure), pragmatic properties and discourse properties. Cognitive Grammar posits that an expression invokes a set of cognitive domains as the basis for its meaning, i.e. as the content to be construed. Therefore, the meaning of a linguistic expression depends on two key notions of ‘cognitive domain’ and ‘construal’. A cognitive domain is a coherent area of conceptualization which provides the conceptual base for the meaning of a linguistic expression. The term construal refers to human manifest ability to conceive and express the same situation or event in alternate ways.
 
Purpose: The present research aims to explore the suffix ‘-ɑr’ within the framework of Cognitive Grammar and Construction Morphology and tries to identify its various constructional schemas and subschemas. Furthermore, it attempts to investigate cognitive domains that underlie the meaning of each subschemas and to analyze the role of construal in formation of each subschemas.
 
Study questions: the study questions are: (1) what are constructional schemas and subschemas of the suffix ‘-ɑr’ and how is hierarchical relationships among them?
(2) What cognitive domain(s) underlie the meaning of each subschema?
(3) Which cognitive processes do play determinant role in formation of these (sub) schemas?
 
Methodology: Empirically, the paper adopts corpus-based method and theoretically, it adopts cognitive and construction-based approach. The data include a corpus of 38 derivational words having the suffix ‘-ɑr’ extracted from the authors own morphological corpus (including more than 10000 complex words) and Farhang-e Zansoo (Keshani, 1993).
 
Analysis: Analysis of research data showed that derivatives of the suffix ‘-ɑr’ belong to different constructional schemas. The schematic-constructional network of the suffix ‘-ɑr’ is represented in Figure 1. As is shown in the figure, ‘-ɑr’ appears in six different subschemas, in five of which the base of derivation is past stem of a verb while in the other the base is a noun. The conceptual base of the five subschemas that have a verb stem as their base is the cognitive domain of ‘processes. The difference among these subschemas is that each subschema profiles a different aspect of the conceptual base. In other words, the difference among them is due to the key notion of ‘profiling’. In the subschema with a noun as the base of derivation, the conceptual base is the cognitive domain of ‘relation’.

Figure 1: The schematic-constructional network of the suffix ‘-ɑr’ in Persian
 
Conclusion: Results of the study show that the suffix ‘-ɑr’, which is added to past stems of verbs or nouns, can appear in six different constructional subschemas to indicate the cognitive domains of process, agenthood, aspect and relation. These subschemas are:
1) <[[x]_(Vpast)i -ɑr]_Nj ↔ [The human agent who perform the process of SEM_i]_j>
2) <[[x]_(Vpast)i -ɑr]_Nj ↔ [The non-human agent (or instrument) which perform the process of SEM_i]_j>
3) <[[x]_(Vpast)i -ɑr]_Nj ↔ [The act of doing the process of SEM_i]_j>
4) <[[x]_(Vpast)i -ɑr]_Nj ↔ [The result of the process of SEM_i]_j>
5) <[[x]_(Vpast)i -ɑr]_Aj ↔ [The property of being affected by the process of SEM_i]_j>
6) <[[x]_Ni -ɑr]_Nj ↔ [The thing closely related to SEM_i]_j>
 
The findings reveal that the cognitive process of ‘construal’ and especially its two aspects of ‘profiling’ and ‘specification’ have a determinant role in the formation of these constructional subschemas.

---

The present study aims to investigate the Ezafe construction in Kurdish (Sorani dialect) based on cognitive grammar. The nature of the methodology of this qualitative research is descriptive-analytic, and the data have been collected via the interview with Kurdish speakers. The authenticity and originality of the collected data have been approved by Kurdish speakers of this linguistic variety. Ezafe marker is considered as the invoker of an intrinsic relationship between two entities having a kind of conceptual proximity.  In a way, one entity is considered as a reference-point based on which the location of the other entity is cognitively determined. In this construction, the landmark plays the role of a reference-point to provide a mental path to the trajector as the target. The findings showed that Ezafe in this dialect is the result of the grammaticalization of a relative pronoun HYA by which the intrinsic-conceptual relation between elements is profiled in a complex sentence. It should be noted that Ezafe marker has developed from the relative pronoun HYA in Old Persian. Accordingly, this relative pronoun, as an Ezafe marker, has been gradually transformed into an element profiling an asymmetric relationship between congruent entities within a single phrase (N Ez N; N Ez adj; N Ez preposition; N Ez nominalized verb) or in a clause (predicative Ezafe and topicalizing Ezafe). Also, the results showed that Ezafe marker indicates this asymmetric relation at different linguistic levels.

1. Introduction
The Cognitive Linguistics approach is considered as a new approach in linguistics (Langacker 1987; Lakoff and Johnson 1980) which tries to explain linguistic phenomena according to cognitive processes without drawing a strict boundary and limitation between linguistic and non-linguistics capabilities. Rejecting the modularity of grammar, it ties to explain syntactic constructions according to the same cognitive processes that are used in non-linguistic phenomena. In harmony with this approach, we try to explain Ezafe construction in Kurdish from a cognitive perspective. In other words, we try to explain the fundamental cognitive processes underlying this frequent and important construction.
The traditional grammarians (Qharib, et al. 1971: 39-45; Moiin 1984; Anvari and Ahmadi-Givi,2011: 134- 138) have described Ezafe as a pure relationship between two entities that may convey a wide variety of meanings. They (ibid) have proposed many functions and senses for Ezafe in Iranian languages. Its numerous functions have been described as Eazfe for possession (ezafeye melki: ketābe ʔali: Ali’s book), Ezafe for expressing the material of a nominal head (ezafeye bayāni: halqeye talā: ring of gold), Ezafe for specifying a noun (ezafeye taxsisi: mize motāleʔe: table for studying), Ezafe for elaborating on the noun (ezafeye tozihi: ŝahre tehrān: city of Tehran) and kinship Ezafe (ezafeye bonovvat: mādar-e ʔisā: mother of Jesus). However, its real semantic and syntactic diversity goes beyond this number. Ezafe linker covers not only a variety of semantic relationships, but also it is used in different grammatical constructions including nominal phrase (nominal collocation: noun Ez noun and nominal periphrasis: noun Ez nominalized verb), prepositional phrase (preposition Ez noun), predicative Ezafe (determiner noun EZ noun copular verb), topicalized structure (topic Ez proposition) and Ezafe for modification by an adjective (noun Ez adjective). This semantic and syntactic diversity leads the formal linguists to consider it as a vacuous element emerging just because of formal computations and structural restrictions (Samiian 1994; Ghomeshi 1997; Samvelian 2005; Karimi 2007; Kahnemuyipour 2014).
The Ezafe construction has been traced back to the (h)ya as a relative pronoun in Old Persian (Moiin 1984: 8-9; 18; Abolqhasemi 2006: 38). The cognitive function of this relative pronoun was for identifying the trajector with another participant in the relative clause. The following example (inscription of Darius at the Naqsh-i-Rostam (DNa)) illustrates the function of hay in Old Persian:
baga vazraka Auramazdā   hya   imām būmim adā,      
the god great Ahuramazda            who  this       earth    created     
hya avam asmānam adā,
who this     sky         created
hya martiyam adā,         hya šiyātim adā  martiyahyā, hya Dārayavaum
who  mankind  created,        who happiness created for mankind, who Darius
 xšāyaθiyam akunauš.
  king made.
Great god is Ahuramazda, who created the earth, who created the sky, who created the man, who created happiness for man, who made Darius king.
It may be argued that the main and primary function of hya is to make a linguistic coordination between two entities that are conceptually parallel and co-equal. However, this relative pronoun undergoes a drastic change in modern Kurdish. It can illustrate any possible relationship between two participants within a single noun phrase. In other words, the Ezafe (as the successor of the hya) is used for demonstrating a relation between two entities that are linked together in a way. Thus, we can say that Ezafe has undergone a kind of subjectification (see Langacker 2009, p.85) through which the relative pronoun (that was to illustrate identification and correspondence between the trajector and a participant in the relative clause) has metamorphosed into a morpheme evoking a schematic model of essential relationship between two participants within a nominal phrase. Through this subjectification, the identification of the sentential trajector with the participant of the relative clause has been turned into any kind of relationship held between two participants in a phrase. Let’s consider the following examples in this case:

1. dæst-i hemən (part-whole)

 hand-Ez Hemn
Hemn’s hand

2. xwardən-i sef (part (patient)-whole (event))

   eating-Ez   apple
eating of apple

3. Halatn-i  Ali (part (agent)-whole (event))

 running-Ez     Ali
Ali’s running

4. gærdænbænd-i tæɬa (material)

  necklace-Ez      gold
necklace of gold

5. ŝār-i bokān (identification)

city-Ez Bokan
city of Bokan

6. raqib-i ʔæli (association)

competitor-Ez ʔæli
Ali’s competitor

7. ktew-i hæsæn (ownership)

book-Ez hæsæn
Hasan’s book

8. bāwk-i hemən (kinship)

father-Ez Hemɵn
Hemɵn’s father

9. hemɵn-i mokryani (nominal collocation)

hemɵn-EZ mokrjani
Hemɵn Mokryani
      The question crossing the author’s mind is that what kind of relationship is evoked by the Ezafe linker that is capable of covering all of these various meninges such as possession, part-whole, participant-event, object-material, nominal collocation, and pure association. It may be discussed that Ezafe evokes a schematic intrinsic relationship that can convey different kinds of semantic association. However, it should be noted that this schematic interconnection is not semantically vacuous. In other words, the linear order of the words in the Ezafe construction cannot be modified, it is because some semantic restrictions dominate this conceptual relationship.
 

1. Literature Review

In this section, we will take a look at previous linguistic studies that have explained Ezafe construction from differet perspective. It is worth mentioning that before this research, some researches have been done in the field of Ezafe construction that have considered this marker as a formal marker and it is explained as the outcome of pure syntactic computations. Some of these researches have been done by Samiian, 1994; Ghomeshi, 1997; samvelian, 2005; Karimi, 2007; kahnemuyipour, 2014. Most previous researches in this area have ignored the polysemic aspect of this construction. Therefore, as far as the authors are aware, no research has been done on the subject of Ezafe construction in Kurdish from a cognitive perspective. Some of the researches done in this regard are those which have been written by Qarib, et al. (1972), Moiin (1985), Abolghasemi (2007), Anvari and Anvari (2011), Razavian, et al. (2016), Amouzadeh and Asmoudeh (2017), Nassajian, et al. (2019).
 

2. Methodology

The nature of the methodology of this qualitative research is descriptive-analytical, and the data have been collected via the interview with Kurdish speakers. The authenticity of the collected data has been approved by Kurdish speakers of this language variety.
 

3. Results

This paper indicated that Ezafe may express different semantic relationship in the level of noun phrase. For example, it can express relations such as ownership, part-whole, identification, and association. However, these different semantic connections are the elaborations of an abstract schematic relation which is evoked by Ezafe. In other words, Ezafe primarily evokes a schematic relation between two elements in which the first one acts as the reference point and the second one functions as the target. The former is used by the conceptualizer to have a cognitive access to the latter.
It was argued that Ezafe is the result of the grammaticalization of a relative pronoun by which the pivot and a participant in the relative clause are matched with one another. The pivot and the relevant participant are characterized by a conceptual contiguity. Accordingly, Ezafe has been gradually turned into an element evoking an asymmetric relationship between contiguous entities within a single phrase (N Ez N; N Ez adje; N Ez preposition; N Ez nominalized verb). However, we showed that this schematic relation is not bound to nominal phrases, but it can play a significant semantic role in the sentential level. As a result, two different constructions such as predicative Ezafe and topicalizing Ezafe were introduced in the paper. In these constructions, the Ezafe mediates between an element and the rest of the sentence. In other words, one of the sentential elements acts as a reference point for the whole sentence. As a result, the conceptualizer gains a mental access to the sentence only via the linguistic element (such as topic) which plays the role of reference point. The common attribute among these various grammatical constructions refers to the intrinsic relationship held between the entities.

---

---

In comparison to other grammatical categories, the verb is considered as an abstract category. Due to hearing limitations, deaf people make mistakes in using words that have abstract meanings; therefore, to examine the verb category, the authors collected about 4500 sentences in the writings of 10 deaf female students in the second and third grades of middle school during an academic year in Kermanshah province. The purpose of this study is to investigate the usage of the two most frequent verbs "be" and "have" in the participants' writings. Following this goal, two questions arose: a) relating these verbs, what kinds of semantic errors are found in the writings of the deaf? b) What are the effects of the semantic and cognitive natures of these two verbs on the extent to which they are used and on their associated errors? The results of this study showed that the selectional errors (using the verbs "be" and " have" instead of each other or instead of other verbs), Argument structure errors, and adding an extra “be” verb error are the high-frequency errors, respectively. This shows that despite the overuse of these verbs, participants are not aware of their semantic distinctions of them. Furthermore, Along with the Cognitive Grammar doctrine, the authors believe that since imperfective verbs’ component states do not change and are constant over time are used more than other verbs. In addition, little content and highly schematic characterization of the verb “be” are the main reasons for the overuse of this verb.
 

1. Introduction

People with mild to profound hearing loss encounter challenges in acquiring language categories due to lack of language exposure. Verbs, which are crucial components of propositions, are particularly difficult to learn because of their abstract nature. The difficulties in verb retrieval observed in individuals with various language disorders suggest that producing and comprehending verbs require greater cognitive efforts compared to other categories.
The current study aims to examine how verbs are utilized in the written language of the deaf. In other words, the authors aim to determine whether this category is correctly used by deaf participants or not. It is important to mention that conducting detailed examinations of verbs in the writings of deaf people necessitates a broad scope of studies. Therefore, this research is being undertaken to meticulously examine the semantic and cognitive characteristics of two particular verbs in Persian through a cognitive linguistics approach offered by Langacker (1987).
Based on the authors' initial observations, "to be" and "to have," which are classified as "imperfective" verbs, exhibited a higher frequency than other verbs. However, simultaneously, they were linked to a greater number of errors. This research is necessary because there has been no study from a cognitive and semantic perspective examining verbs in the deaf language production.
 Research Question:

1. What types of semantic errors are evident in association with the verbs "to be" and "to have" in the written language of deaf individuals?
2.  what influences do the semantic and cognitive characteristics of these two verbs have on the frequency of their usage and any related errors?

2. Literature Review

Hassani Jalilian (2020) demonstrated that the deaf students utilized less- prototype verbs to a large extent. The abstract nature of the verb category, the semantic characteristics of less prototype verbs, the limited language exposure, and the different social experiences of the individuals were cited as possible explanations for this phenomenon.
Gheitury and Choubsaz (2017) examined the semantic and linguistic knowledge of 12 deaf students through pictorial and written tasks. The study aimed to assess their knowledge of argument structure, which involved determining the number of arguments each verb required. The result revealed that the deaf participants had a satisfactory comprehension of the semantic properties of verbs as well as the number of arguments each verb took.
In their research, Ormel et al. (2020) explored the difference in semantic knowledge between bilingual deaf children and their hearing peers at different ages. They used images, writing, and sign language to evaluate semantic categorization at the exemplar and subcategory level.  The result of their research indicated that, on average, the semantic knowledge of the deaf children was not well-developed compared to their hearing peers.

3. Methodology

3.1 Participants
To conduct the present study, interviews were carried out with 23 deaf and hard of hearing students as well as their parents and their teachers. Then, 10 deaf individuals who were similar in terms of hearing, physical, cultural, and economic factors were chosen. The average age of the selected participants was 17.2 years (SD = 1.13). A control group of hearing peers matched for age and education was also selected. The deaf group were in the second and third grades of middle school. All deaf participants grew up in monolingual (Persian) families and had not used any specific sign language during their childhood. Instead, they used home signs before starting school.

3.2 Data Collection
The data collection process lasted approximately 7 months during an academic year. The data were collected through weekly meetings (each lasting approximately 20-30 minutes). The participants were asked to write compositions about abstract, less abstract, and concrete subjects. The students were given pictures for less abstract and concrete subjects. In addition, when they needed assistance, their teacher would provide explanations and ensure they clearly understood the concept. Overall, 4504 sentences were collected from the deaf group for further analysis.

4. Results

The findings showed that deaf people faced challenges in using the verb category. In the deaf writings verb diversity was significantly low compared to their hearing group. Selectional errors (using the verbs "be" and " have" instead of each other or instead of other verbs), Argument structure errors and adding an extra “be” verb errors were the high frequency errors, respectively. The errors demonstrated that due to the verbs’ abstract nature and requiring more cognitive load, they cannot be acquired naturally by the deaf individuals. 
Another argument that can be put forward for relying heavily on these two verbs is related to the semantic distinctions between the perfective and the imperfective verbs as proposed in Cognitive Grammar. Imperfective verbs, which have a lower semantic content compared to other types of verbs, are easier for the deaf to learn and use.
Based on the entrenchment phenomenon proposed by Langacker (2008), the frequent use of "be" and "have" led to their status as active linguistic units. This means that when deaf individuals do not have the appropriate verbs in mind, they might substitute these verbs instead. It should be acknowledged that the different ways of using verbs by deaf individuals do not imply that they do not comprehend them, since as stated in the Cognitive Grammar, comprehending verbs requires understanding relationships and tracking time, which is a natural cognitive ability among human beings. That is, a deaf person knows the concept of verbs such as "to break" or "to clean," but due to the lack of linguistic experience, s/he cannot learn the corresponding linguistic labels to use them in necessary situations.
 

---

This paper is an attempt to explore the prepositions used to encode the agent in Persian passive constructions. According to Givón (1983) the main function of the passive constructions is to defocus the agent. As the general trend in Persian is to omit this defocused agent, we will stylistically divide the passive constructions into two main sub branches, namely unmarked and marked passives according to whether or not the agent is included. Having analyzed the prepositions using in the passive construction (e.g. t{v{ssot-e, be d{st-e, bA, {z suy-e, {z t{r{f-e, and be v{sile-y-e) a complete semantic characterization of the type of agent that normally occurs with these prepositions will be provided. The results indicate that, in contrast to some linguists who treated the prepositions in the passive construction as stylistic forms, they are not always interchangeable; in fact they are meaningful which have a prototype for the type of agent that co-occurs with them.

---

According to the extended projection principle (EPP), it is claimed that all sentences require subjects. In line with such assumption, it is believed that some languages, including Persian, are categorized as pro-drop languages and their subject is encoded by the verbal inflection. In fact, the subject position is an empty category designated by pro (small pro). Therefore, AGR (verbal inflection) has a purely syntactic function. However, in cognitive grammar, AGR is treated as a symbolic assembly profiling a process whose only independent contribution to the meaning of the verb with which it combines, is the person and number specification of the processual AGR. Considering AGR as a meaningful category, its trajector can correspond with the trajector of the processual component (verb) which is left unexpressed. In contrast to the assumption that AGR is redundant, it shown that the subject (the trajector of the processual component) is not dropped; it does exist but is highly schematic. It is elaborated (becomes specific) by trajector of AGR through correspondence. In other words, Persian speakers conceptualize the subject by only one source of information; that is the trajector of AGR.