Hostname: page-component-7479d7b7d-767nl Total loading time: 0 Render date: 2024-07-08T23:22:19.193Z Has data issue: false hasContentIssue false

Comparison constructions in two Northern Talyshi dialects

Published online by Cambridge University Press:  05 December 2023

Hakob Avchyan*
Affiliation:
Department of Iranian Studies, Yerevan State University, Yerevan, Armenia
Rights & Permissions [Opens in a new window]

Abstract

This article examines the comparison constructions in two Northern Talyshi varieties: in Anbarāni, used in the Islamic Republic of Iran; and the Northern Talyshi dialects spoken in the Republic of Azerbaijan. These constructions have been poorly studied in previous research dealing with this North-western Iranian language and this article aims to fill that gap. In contrast with a number of Western Iranian languages, Northern Talyshi (and the Talyshi language in general) does not have special morphological markers for expressing the degrees of comparison. The comparative grades are marked syntactically using various adpositions and function words. Having long been under the influence of neighbouring languages such as Persian and Azerbaijani, both Anbarāni and the Talyshi dialects spoken in the present-day Republic of Azerbaijan have been affected by these languages at some level in the ways to make comparison.

Type
Article
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The Royal Asiatic Society

Introduction

TalyshiFootnote 1 is a North-western Iranian language that is spoken in the Islamic Republic of Iran and the Republic of Azerbaijan. The land of Talysh is located in the south-western region of the Caspian Sea and generally stretches from the south-east to the north for more than 150 kilometres.Footnote 2 In the Republic of Azerbaijan, most Talyshis live in a region extending from the western bank of the Vilaj River in the north to the Iranian border in the south and from the Caspian Sea in the east to the Iranian border in the west—that is, in the five districts of Āstārā, Lankarān, Lerik, Māsālly, and Yardymly that include over 350 Talysh villages and towns.Footnote 3 As for the Islamic Republic of Iran, Talysh people compactly dwell mainly in the provinces of Gilān (in the counties of Āstārā, Tālesh, Rezvānshahr, Māsāl, and Fuman) and Ardabil (mainly in Namin county as well as in Khālkhāl and Bileh Savār). The Talysh language includes a wide range of dialects, which are traditionally divided into three main groups, namely Northern, Central, and Southern, based on phonological, grammatical, and lexical factors.Footnote 4 Northern Talyshi is spoken almost exclusively in the Republic of Azerbaijan, but it is also used in the neighbouring parts of the Islamic Republic of Iran, in the Anbarān district of Ardabil, the only area in the Islamic Republic of Iran where the Northern dialect of Talyshi dominates in the social domain in more than one neighbouring village.Footnote 5 Thus, Northern Talyshi in turn splits up into several clusters; Talyshi spoken in the Republic of Azerbaijan is divided into four main groups: dialects of Lankarān, Āstārā, Lerik, and Māsālly.Footnote 6 There are a number of insignificant phonological, morphological, and lexical differences between these dialects.Footnote 7 Anbarāni is the southernmost dialect of the Northern group and, as already mentioned, is a Northern Talyshi variety found in Iran. Anbarāni and the Northern Talyshi dialects of the Republic of Azerbaijan are completely mutually intelligible but, since they have been divided between two countries and thus have long been under the strong influence of Persian and Azerbaijani languages,Footnote 8 the phonological, morphosyntactic, and lexical dissimilarities between Anbarāni and Azerbaijani Talyshi are quite considerable, including those of the comparison constructions, which will be further discussed.

The aim of this article is the synchronic comparative description of positive, comparative, superlative, equative, and similative constructions in Anbarāni and the Talyshi dialects of Azerbaijan. Comparison is an act of examining two or more items to reveal the similarities and differences between them. Stassen defines a comparative construction: ‘a construction in a natural language counts as a comparative construction (and will therefore be taken into account in the typology) if that construction has the semantic function of assigning a graded (i.e., non-identical) position on a predicative scale to two (possibly complex) objects.’Footnote 9 Traditionally, in grammar, four degrees of comparison of the adjective are differentiated and the labels for these morphological forms of the adjective are also applied to the whole comparison construction in which they are used: positive degree, equative degree, comparative degree, and superlative degree.Footnote 10

To describe the comparison constructions in the two varieties of Northern Talyshi, this article mainly adopts the following terminology used by Dixon:

compareethat which is being compared,

standard of comparison—what the comparee is being compared against,

parameter of comparison—the property of comparison,

index of comparison—marks the degree of presence or absence of a property in the comparee, and

mark of the grammatical function of the standard—marker of the grammatical function of the standard.Footnote 11, Footnote 12

The data

The data on Anbarāni (AnbTal) upon which this article is based were recorded during fieldwork conducted between 2015 and 2017 in the former village of AminjānFootnote 13 (Ardabil Province, Namin County, Anbarān District). Example sentences were drawn from a corpus including 30 texts (folktales, anecdotes, translations from Persian, and personal narratives) and a sentence list consisting of more than 300 sentences. As for the Northern Talyshi dialects spoken in the Republic of Azerbaijan (AzTal), due to the impossibility of carrying out fieldwork in the area in which these dialects are spoken, data gathered through modern technologies for remote communication, as well as a number of published language materials such as fairy tales miscellanea and works of Talyshi writers and also newspapers and texts from Talyshi-language news portals, were taken into account.

Background of the informants

My main informant from Anbarān was Sinā Moradzādeh—a 20-year-old student at the University of Mohaghegh Ardabili, who had perfect command over both his mother tongue and Persian, and helped me to also write down and transcribe the language material provided by other informants; Safā Ghamkhāri—a 55-year-old bilingual man (Talyshi and Persian) working as an independent contractor (peymānkār) and having a high-school education; Ruhangiz Ehsāni—an 81-year-old trilingual woman (Talyshi, Persian, and Turkic) without a school education; and Shaghāyegh Ghoreyshi—a 19-year-old bilingual girl (Talyshi and Persian) and a student at the Ardabil University of Medical Sciences.

Two of my informants of AzTal with whom I worked through video technologies for remote communication preferred to remain anonymous. Both of them are male, bilingual, and fluent in Talyshi and Azerbaijani. One has a high-school education and the second has a master's degree and currently teaches in a university.

Northern Talyshi: morphosyntactic background

Talyshi is both genealogically and structurally very close to the highly diversified Tati languages spoken in the Iranian plateau with which it is unified into one Tatic group.Footnote 14 Being one of the North-western Iranian languages that descended from an Old Iranian dialect (or dialects) that was already differentiated from Old Persian (a South-western Iranian language, the ancestor of Modern Persian) in its earliest attested form in the sixth century BC, one can assume a time-depth of some 2,800 years in the North-western Iranian/South-western Iranian split, thus making Talyshi, in general terms, much further apart from Persian than English is from Swedish or French from Romanian.Footnote 15

Here, I briefly introduce the salient morphosyntactic features of Northern Talyshi that will help the readers to have a better understanding of the main topic of the article and clarify the individual examples.

Nouns in Northern Talyshi possess the following grammatical categories: number, case, and definiteness–indefiniteness. The grammatical descriptions of Northern Talyshi are at variance with the number of cases in this language. Some authors believe that Northern Talyshi has two cases: direct and oblique,Footnote 16 while Boris Miller distinguishes three cases for Northern Talyshi: direct, oblique, and ablative cases.Footnote 17 Further, Schulze includes the vocative amongst the cases in Northern Talyshi.Footnote 18 This article follows the two-case system to describe the comparison constructions in Northern Talyshi. Both in AnbTal and AzTal, nouns in direct case have zero ending. The oblique case ending is in AnbTal and -i in AzTal.Footnote 19 Whereas in AzTal, in most dialects, -i usually changes into -y when attached to vowel-ending nouns (e.g. kay ba ‘the house's door’ and kinay moa ‘the girl's mother’), in AnbTal, in this case, the distinction between the direct and oblique cases vanishes. One of the main functions of the oblique case in Northern Talyshi is the expression of genitive—that is, marking a person or thing that possesses someone or something (see examples (16), (32), (33), (42), (44), and (46)). It also serves to mark the definite direct object in accusative constructions (as seen in examples (1) and (2)) and the agent in ergative constructions (as seen in examples (6) and (7)):

Due to the limited number of cases, the syntactic relationship between words in sentences in Northern Talyshi is conveyed using helper words, including prepositions (čə ‘from, of’, ba ‘to’, bo ‘for’, be ‘without’, de/da ‘with’, etc.) and postpositions (ku ‘from’, sa ‘on’, ada ‘in’ (AzTal)/anda ‘in, with’ (AnbTal), etc.). Most of these adpositions are used with nouns in the oblique caseFootnote 20 (see examples (12) and (13)). The core postpositions are encliticsFootnote 21 while the prepositions are used independently. Note that, while all prepositions never attach to nouns, they can be used either proclitically or independently with personal, demonstrative, and reflexive pronouns, e.g. bamə or ba mə ‘to me’, bašma or ba šəma ‘to you’ (PL.), bay or ba ay ‘to it, to him/her’ čəm or čə əm ‘of this’, ča or čə a ‘of that’, day or de ay ‘with it, with him/her’, dešta or de əšta ‘with his/her own…’, bošta or bo əšta ‘for his/her own…’, etc.

The Northern Talyshi personal pronouns differentiate three persons in the singular and plural, which are as follows: 1SG. āz, 2SG. , 3SG. av, 1PL. ama, 2PL. šəma, 3PL. avon (AzTal)/avun (AnbTal). These are the forms in the direct case. The first and third singulars also have oblique forms (1SG. and 3SG. ay in AzTal, 1SG. mān/mānə and 3SG. avə in AnbTal). The direct and oblique forms for the rest of the persons and numbers are the same.

AzTal has also developed secondary accusative forms that are not found in AnbTal. AzTal secondary accusative forms of personal pronouns are: 1SG. məni, 2SG. təni, 3SG. avi, 1PL. amani, 2PL. šəmani, 3PL. avoni. They act only as direct objects in accusative structures:

The direct forms of personal pronouns have a wide range of syntactic roles, whereas the first and third singular oblique forms mainly function as agents in ergative constructions (see examples (29), (30), and (60)). Northern Talyshi also has possessive pronouns, namely 1SG. čəmān (AnbTal)/čəmə, 2SG. əštə, 3SG. čavə (AnbTal)/čay (AzTal), 1PL. čama, 2PL. šəma, 3PL. čavun (AnbTal)/čavon (AzTal). These are analytic forms consisting of the preposition čə and personal pronouns (the oblique forms for the first and third singulars).Footnote 22 The possessive pronouns always precede the nouns modified by them, are invariable, do not express the grammatical number of the nouns they qualify, and are only used as attributes (see examples (10), (11), (17), (20), (38), (39), (44), (45), (49), (55), and (61)).

In Northern Talyshi, the verb stands out among the parts of speech for its richness and variety of grammatical categories and forms; it has categories of mood, tense, person, number, and voice. The salient characteristics of the NT verbal system include the opposition between the ‘present’ and ‘past’ stems,Footnote 23 the contrastive roles of suffixes and floating clitics,Footnote 24 and tense-sensitive alignment.Footnote 25

The morphosyntax of Northern Talyshi is characterised by a complicated split system based on the North-west Iranian type of accusativity/ergativity dichotomy.Footnote 26, Footnote 27 The term ‘ergativity’ is broadly used to describe a ‘grammatical pattern in which the subject of intransitive clause [S] is treated in the same way as the object of a transitive clause [O], and differently from a transitive subject [A]’.Footnote 28 Northern Talyshi has what has been called split ergativity—that is, some constructions use the ergative syntax and morphology, while others display an accusative pattern. The ergative construction is limited to transitive past-perfective environments (i.e. clauses in which the predicate is in one of the following tenses: past simple, present perfect, or past perfect). All other intransitive or transitive non-past-perfective forms trigger accusative marking. In this case, the subjects of intransitive verbs and the subjects of transitive verbs form a morphosyntactic unity in the way that both are marked by the same case (direct case). The direct object of the transitive verb, however, is distinguished from the subject by the oblique case (which is traditionally called ‘accusative’)Footnote 29:

In the NT ergative constructions, the agent [A] of the verb is in the oblique case, the direct object [O] is in the direct case, and the number and person of the agent are expressed by enclitic pronounsFootnote 30:

The placement of the enclitic pronouns displays great variety in NT and can be detached from the verb and attached to any component part of a sentence except the agent itself (see example (30), in which the enclitic pronoun is attached to the direct object, or (60), in which it is attached to the word ğadar; however, in example (61), it remains in the verb).

Adjectives in Northern Talyshi: the positive degree of comparison

Adjectives in the Northern Talyshi dialects lack the grammatical categories of case, number, and gender, as well as the category of definiteness or indefiniteness. Northern Talyshi adjectives can be divided into two groups in terms of their forms: simple and complex adjectives. Simple adjectives are root words in their own right and do not have prefixes or suffixes, such as yol ‘big’, māhbub (AnbTal)/rečin (AzTal) ‘beautiful’, tərš ‘sour’, sərāf ‘fat’, etc. Complex adjectives are split into two groups in their turn: derived and compound adjectives. Derived adjectives are formed chiefly from nouns and other adjectives by adding various suffixes and prefixes. The main affixes to form derived adjectives in both varieties of Northern Talyshi are as follows:

Compound adjectives are formed by combining different adjectives, nouns, verb stems, numerals, and pronouns as follows:

Attributive adjectives describe a characteristic or an attribute of the noun or the noun phrase that they modify. In the Northern Talyshi dialects, attributive adjectives precede their head and are normally marked by the attributive marker -a, such as māhbuba kina (AnbTal)/rečina kina (AzTal) ‘beautiful girl’, šina āngə ‘sweet grapes’, tāta āš ‘hot soup’, yola ka ‘big house’, etc.Footnote 31

Adjectives with a predicative sense, hence the positive construction (e.g. A is big) has the construction of noun + adjective + copula:

In the positive degree of the comparative construction, attributive and predicative adjectives can modify the same noun:

In AnbTal and AzTal, the positive degree of the comparative construction has exactly the same structure.

Comparative constructions

Adjectival comparative constructions

The Northern Talyshi comparative construction of superiority (e.g. A is bigger than B) is mono-clausal—the comparee and Standard are expressed by noun phrases whose head is a noun or a personal/demonstrative pronoun, with the parameter as a copula complement. Both AnbTal and AzTal normally have zero (ø) in the index slot. As for the mark of the standard of comparison, all Northern Talyshi dialects use a number of postpositions and prepositions or their combination.

In both AnbTal and AzTal, the main postposition functioning as the mark of the standard in the comparative construction is sa Footnote 32 ‘on’:

Another postposition used in both the dialects for marking the standard is the postposition ku Footnote 33 ‘from’. In AzTal, the standard marked by it is often preceded by the preposition čə Footnote 34:

In AzTal, there are two additional ways of marking the standard that are not found in Anbarāni. The first is the occasional use of the postposition ada Footnote 35 ‘in’ as the mark:

The second option for marking the standard, which is more commonly used, can be viewed as another variation of the construction with the mark sa that was previously discussed, but it also involves the use of the compound preposition basa ‘on, onto’ before the standard:

It is interesting that Anbarāni normally does not use an index but, when the standard and its mark are omitted in an appropriate discourse context, the parameter obligatorily takes the index -tar,Footnote 37 borrowed from the Persian languageFootnote 38:

Examples (18) and (19) were drawn from two different personal narratives of two of the informants. In the first narrative, the informant speaks of roses and says that he does not like them. Then he states that he loves daisies a lot: ‘These flowers are more beautiful.’ As for example (19), it was stated in response to my statement: ‘Surely, that big room is yours, and this small one is your grandson's’ when the other informant was showing his house and describing it.

In analogical appropriate discourse contexts in AzTal as well, the standard can be omitted, and the words haniyan ‘even more’ and tikayan ‘a little more’ are used as indexes marking the parameter:

As previously mentioned, in both AnbTal and AzTal, the comparative construction normally lacks the index, but the words ve/vey (AzTal), vəyi, xayli (AnbTal) ‘much, a lot’ are found to be used sporadically in the constructions under discussion:

Adverbial comparative constructions

The examples provided thus far have dealt with the adjectives as copula complements—that is, the parameters of these comparison constructions describe the characteristic of the subject, thus the comparee is the copula subject of the clause; in other words, it is not morphologically marked because, in both AnbTal and AzTal, copular clauses always have an accusative structure. While adjectives modify nouns or noun phrases, adverbs modify verbs; consequently, gradable comparisons can be classified into two main types: adjectival comparative constructions and adverbial comparative constructions (e.g. A runs faster than B). In Northern Talyshi, the number of manner adverbs is extremely small and mainly adjectives are used in adverbial function; therefore, in adverbial comparative constructions, the parameter always acts as a manner adverbial, the comparee can have various grammatical functions, while the standard is seemingly always the indirect object of comparison. As in the adjectival, adverbial comparative constructions usually do not have index either and the same adpositions are used as the mark.

In example (24), the comparee toləši functions as the direct object, the standard tərki is the indirect object of comparison, and the adjective čok has a manner adverbial function and is used as the parameter. The construction lacks the index. In AzTal, this kind of adverbial comparative construction has exactly the same structure (not taking into account the dialectical dissimilarities not affecting the comparative constructions) with the comparee, standard, and parameter with the same grammatical functions:

In example (26), (27), and (28), the parameter describes the characteristic of the action carried out by the comparee, which functions as the subject, in comparison with the standard:

In contrast with adjectival comparative constructions in which the comparee with the grammatical function of the subject is always morphologically unmarked, due to the fact that Northern Talyshi shows a split ergative pattern, the COMPAREE of adverbial comparative constructions can act as the agent of the clause and consequently will be marked as such:

Thus, summing up all the aforementioned and taking into account the criteria for identifying comparative types suggested by scholars who have discussed the topic, the Northern Talyshi comparative constructions can be categorised. Dixon defines the types of comparative constructions according to morphosyntactic parameters, as Treis claims that ‘[i]mportant features for the categorisation seems to be the mono/bi-clausality of the construction, the syntactic functions of the primary components of a comparison scheme and the degree of grammaticalisation of the construction’.Footnote 41 Thus, according to Dixon's typology,Footnote 42 Northern Talyshi has an A1 type of comparative construction—that is, the comparee is the copula subject, the standard is an oblique NP, the parameter is the head of the copula complement, the mark marks the oblique NP, and the index modifies the copula complement, although both AnbTal and AzTal comparative constructions usually lack it. Dixon's A1 type only applies to adjectival comparative constructions. The adverbial comparative constructions are defined as one of the ‘other schemes of comparison’.Footnote 43

Stassen identifies the comparative types preponderantly by the type of mark of the standard.Footnote 44 According to Stassen's sample, a language can have more than one option for marking the standard; thus, there can be several options for the comparative-type choice. The type that is more widely used in the language is the primary comparative construction and the others are consequently called secondary comparatives.Footnote 45 Northern Talyshi has what are called fixed-case comparatives. These comparatives are in turn divided into exceed comparatives and locational/adverbial comparatives. Northern Talyshi comparative constructions, as displayed through the provided examples, are locational/adverbial—that is, the standard NP is expressed by a phrase that is in an adverbial relation to the parameter. These constructions comprise three subtypes, two of which are typical for Northern Talyshi. Both AnbTal and AzTal have type 3 (the locative comparative) as their primary comparative constructions—that is, ‘the standard NP is invariably encoded as a constituent of an adverbial phrase which is marked by an element that indicates spatial or non-spatial contact’.Footnote 46 The postposition sa ‘on’ is used as the mark of the standard in both the Northern Talyshi varieties. In AzTal, the standard marked by sa can be additionally preceded by the compound preposition basa ‘on, onto’.

Stassen's type 1 (the separative comparative) is the secondary comparative construction found in Northern Talyshi. The postposition ku ‘from’ functions as the mark to modify the standard; it is used more frequently in AzTal and is often preceded by the preposition čə ‘from’.

None of the Northern Talyshi dialects seems to have a grammaticalised or conventionalised comparative construction of inferiority. Instead, it can be expressed through the negation of equative comparative constructions (see ‘Equative constructions’, below):

Superlative constructions

The superlative expresses the highest degree of a quality that someone or something possesses—that is, the comparee has a parameter to a higher degree than each individual entity in a group of more than two.Footnote 47 Several types of superlative constructions (e.g. A is the biggest) are used in Northern Talyshi. The first type that is found in both AnbTal and AzTal is based on the comparative construction with the standard expressed by the pronoun hama (AnbTal)/hamma/hammay (AzTal) ‘all, everybody’, which is modified by the mark sa as in the comparative constructions:

In AnbTal, this type of superlative is based only on the locative comparative constructions with the standard marked by sa ‘on’. In AzTal, the standard of superlative constructions is hardly ever marked by the postpositions ku ‘from’ and ada ‘in’,Footnote 49 but superlatives with the mark basa… sa are not rare:

In Northern Talyshi, this type of superlative construction can be paraphrased; thus, the copula complement becomes the standard of the construction, the pronoun hama/hamma precedes the standard (naturally, if it was singular, it becomes plural), and then the mark sa modifies it. In this kind of construction, the parameter acts as the copula complement, cf. examples (38) and (39); (40) and (41):

The second type of superlative construction found in Northern Talyshi is the result of the influence of neighbouring languages on Northern Talyshi and is characterised by the obligatory use of a dedicated degree marker, which accompanies the parameter.Footnote 50 In AzTal, this is the most common type to form superlatives and is found more often than the first type described above. This type of superlative construction in AzTal is formed by adding the word an ‘most’, a borrowing from Azerbaijani,Footnote 51 which precedes the parameter:

It is interesting that this type of superlative construction is not described by Boris MillerFootnote 52 nor is it found in any other work on Azerbaijani Talyshi.Footnote 53 This fact likely shows that it is a recent borrowing that is rapidly spreading. It seems that, in AzTal, the degree marker an modifies the parameter when it is an attributive adjective as in examples (42) and (43) and does not act as the mark with predicative adjectives or the adjectives functioning as manner adverbs. In this case, the pronoun hamma is used, as in examples (38) and (39).

This type of superlative is rarely found in AnbTal and is formed by the suffix -tarin borrowed from Persian.Footnote 54 It appears twice in my text corpus in the following sentences:

It is also worth noting that, as we can judge from examples (44) and (45), the parameter modified by -tarin does not take the attributive marker -a, which, as has been discussed in ‘Adjectives in Northern Talyshi’ above, normally marks the attributive adjectives preceding their head. This is in all probability triggered by the fact that the Persian superlative adjectives suffixed by -tarin precede the head and, if the head is singular, they do not take ezafe.

The third means of expressing the idea of the superlative is the usage of the intensifying adverb lāp ‘very’, which modifies the parameter and precedes it. This is the only type of superlative construction described by Boris Miller. Examples of this type are illustrated in examples (46) to (51):

AzTal has a superlative-like form that is expressed by means of the reduplication of adjectives. In these constructions, the first copy of the reduplicated adjective is used substantively, is in the oblique case, and is marked by the postposition sa. The second component is unmarked and functions as a copula complement:

The use of these constructions seems to be quite limited and the parameter always has the grammatical role of a copula complement. Besides, these forms express not the sense of comparison, but the highest degree of a quality; thus, they should not be considered real superlatives. In my corpus of Anbarāni, no example of this construction could be identified.

In summary, in the Northern Talyshi dialects, three of the five types of superlative constructions suggested by GorsheninFootnote 55 are found:

  1. 1) Type A = Absolute comparison superlative that is ‘based on a comparative construction with the standard expressed by a universal quantifier all, everybody, everything as head of the phrase or as modifier’.Footnote 56 In both AnbTal and AzTal, the pronoun hama/hamma(y) functions as such a quantifier that is mainly modified by the mark sa (AnbTal, AzTal), basa… sa (AzTal), and also sporadically by ku and ada in AzTal.

  2. 2) Type DEG = Conventionalised degree superlative ‘is characterised by the obligatory use of a dedicated degree marker, which accompanies the parameter and can either be free or bound’.Footnote 57 This type in both AzTal, where it appears quite frequently, and in AnbTal, where it rarely occurs, results from the influences of Azerbaijani and Persian, respectively. In AzTal, the degree marker an modifies the parameter, which is in turn linked to its head by the vowel -a. In AnbTal, the superlative suffix -tarin, borrowed from Persian, marking the parameter in the superlative construction does not bear the attributive marker -a (examples (44) and (45)) in contrast with the superlatives formed by the means hama: nəğəltarin gol, but hamasa nəğəla gol ‘the deepest lake’, mehrabuntarin žen, but hamasa mehrabuna žen ‘the kindest woman’.

  3. 3) Type INT = Intensifier superlative, which ‘includes an overt degree marker, which is, however, not restricted to the expression of superlativity but has a general intensifying function “very,” “by far,” etc.’.Footnote 58 The main degree marker used in this type of superlative construction both in AnbTal and in AzTal is the adverb lāp ‘very’, which marks the parameter in all its grammatical functions, including a copula complement as in examples (51), an attribute as in examples (46), (47), (49), and (50), and a manner adverb as in example (48).

Equative and similative constructions

Equative and similative constructions share close semantic and formal similarities; thus, they are usually discussed together in typological works. Equatives express sameness of degree or extent, while similatives express sameness or identity of manner.

Equative constructions

The equative constructions (e.g. A is as big as B) in Northern Talyshi are very similar to the adjectival comparative constructions with almost the same components (A—comparee, B—standard, as—standard-marker (= mark), big—parameter) and structure. Both AnbTal and AzTal lack the degree-marker (as). In AnbTal, the words kimi, taki/takina, bata ‘like’ are used as the standard-marker. Kimi, which is borrowed from Turkic, and taki/takina are postposed, while bata precedes the standard. The standard is always in the oblique case:

In AzTal, the main word acting as the standard-marker is bana ‘like’, which precedes the standard in the oblique case:

In all the provided examples (54) to (58), the parameter is the copula complement; thus, as in the adjectival comparative constructions, the comparee is unmarked, and the standard is always in the direct case marked by the standard-marker.

Northern Talyshi also has what is called quantitative equative constructionsFootnote 59 that express quantity (e.g. A has as many books as B). Both AnbTal and AzTal use the TurkicFootnote 60 loanword ğadar ‘as/as much as’. It is used with the standard in the oblique case or with possessive pronouns:

As seen from examples (59) to (61), the peculiarity of the quantitative equative is that it is the only comparison construction in Northern Talyshi wherein the parameter is expressed by nouns. Besides, unlike the qualitative equatives, the comparee acting as subject, depending on the used verb, its transitivity/intransitivity, and the tense, can be either in the direct case and thus unmarked or in the oblique case and marked as such.

Many languages morphologically differentiate between specific and generic equatives. Rett describes the generic equative as ‘an adjectival equation construction in which the subordinate clause (the one introducing the standard) has a generic or habitual aspect’.Footnote 61 Haspelmath and Buchholz argue that often the standard in such comparisons is a class but possesses the property in question to a highly salient degree, and the comparison must not be taken too literally, such as ‘white as snow’, ‘heavy as lead’, ‘dark as night’, ‘cunning as a fox’, etc.Footnote 62 In Northern Talyshi, no difference exists between specific (examples (54) to (58)) and generic constructions; both have exactly the same structure and components, and the same words are used as the standard-marker:

In sum, Haspelmath and Buchholz classify the equative constructions found in European languages into three main types: 1) relative-based equative constructions, 2) constructions primarily characterised by a parameter marker, and 3) constructions primarily characterised by a standard marker.Footnote 63 Northern Talyshi clearly belongs to the third group that completely lacks a degree-marker and is only characterised by the connective words bata, taki/takina, kimi (AnbTal), bana (AzTal) ‘like’, and ğadar (AnbTal, AzTal) ‘as/as much as’. All these words combined with the standard place it in the oblique case. In more recent work, Haspelmath et al. propose to distinguish six primary types of equative constructions.Footnote 64 Both the Northern Talyshi varieties belong to type 1: only the equative standard marker (‘Kim is tall like + Pat’.), which is the most common of all types.

Similative constructions

Similatives (e.g. A runs like a rabbit) are comparative constructions with respect to the manner in which an action is fulfilled. Equative constructions express the identity, which is a simple one-dimensional notion, whereas manner is a complex multifaceted notion. Hence, in general, only equatives really express equality, while similatives tend to express similarity.Footnote 65 Similative constructions in AnbTal and AzTal are similar to quantitative equative constructions in terms of their structure; in both constructions, the comparee is the subject of the clause, and the standard is marked by a standard-marker functioning as a manner adverbial that modifies the predicative verb:

The only structural difference between examples (66) and (67) is that the quantitative equative construction also contains a parameter whereas, in examples (67), the parameter is not overtly expressed.

In similative constructions in both AnbTal and AzTal, the same connective words are used as standard-markers as in normal, namely qualitative (≠quantitative) equative constructions:

Conclusion

The main features of comparison constructions in the Northern Talyshi dialects spoken in the Islamic Republic of Iran and the Republic of Azerbaijan can be summarised as follows. Northern Talyshi does not have any special morphological means for marking the comparative grade; hence, both AzTal and AnbTal rely on syntactic frames to express the comparative, superlative, equative, and similative degrees. The comparative construction of superiority mainly comprises four components: comparee, standard, parameter, and mark. Both AnbTal and AzTal comparative constructions almost always lack the fifth component, namely the index or degree-marker, but it can be sporadically expressed using manner adverbs such as ve/vey (AzTal), vəyi (AnbTal), and xayli ‘much, many, a lot’. The parameter of comparison is a gradable adjective. The main postposition acting as the mark of the standard is sa ‘on’ in AnbTal and AzTal. The standard can be also marked by the postposition ku ‘from’ in both the Northern Talyshi varieties. The postposition ada ‘in’ functioning as the mark appears rarely and is found only in AzTal. All postpositions in AzTal can be combined with the preposition čə ‘from’; besides, in this variety, the circumposition basa… sa modifying the standard is quite frequent. On this level of comparison, the influence of neighbouring languages is found in AnbTal when, in an appropriate discourse context, -tar, borrowed from Persian, is used as the index.

The main type of superlative construction in Northern Talyshi is based on the comparative construction. The standard of the comparison here is the pronoun hama (AnbTal), hamma/hammay (AzTal) ‘all, everybody’, which is marked by the same adpositions as in the comparative construction. The second type of superlative is characterised by the obligatory use of a dedicated degree-marker encoding the standard, which is an ‘the most’—an Azerbaijani loanword—in AzTal and appears very frequently whereas, in AnbTal, the Persian-borrowed suffix -tarin is very rarely used. There are also intensifier superlatives that contain an overt degree marker having a general intensifying function. In AnbTal and AzTal, the adverb lāp ‘very’ functions as such a marker.

The specific and generic equatives in Northern Talyshi are structurally very similar to adjectival comparative constructions. They include exactly the same components but, instead of the adpositions functioning as the mark of the standard, the words kimi, taki/takina/bata (AnbTal), bana ‘like’ are used as the standard-marker. In quantitative equative constructions, both the Northern Talyshi dialects utilise the Turkic-borrowed word ğadar of Arabic origin as the standard-marker. The similative constructions are structurally identical to quantitative equatives; however, the parameter is expressed covertly, and the standard is modified by the markers used in specific equatives.

Acknowledgements

This article was written during my research stay at the Institute of Iranian Studies of the Austrian Academy of Sciences in the frames of the Joint Excellence in Science and Humanities (JESH) programme. I take this opportunity to convey my heartfelt thanks to the invitation of Prof. Dr. Florian Schwarz. I would also like to express my sincere gratitude to Sinā Moradzādeh for his immense help with writing down and transcribing the data on Anbarāni, as well as for offering the needed information and explanations. I am especially thankful to my colleague and friend Dr. Artyom Tonoyan for reading this article and his valuable comments and suggestions.

Conflicts of interest

None.

Abbreviations

Appendix

In this Appendix, two versions of the same text are provided. The AzTal version, drawn from the ‘Manual of Talyshi’,Footnote 66 has been prepared by the authors and checked and edited by a native speaker who preferred to stay anonymous. The Anbarāni version has been translated by Sinā Moradzādeh, a student from Anbarān with perfect command of his mother tongue, from Persian.

Azerbaijani Talyshi (AzTal)

Vist sorisa veye ki čəmə moa čə kišvari hammaysa yola universitetada ko kardeda. Av ingilisa zəvoni miallimaye iyan əm zəvoni vey čok zəneda. Moa hežo bašta talabon voteda ki ingilisa zəvon čə dənyo an ganjina zəvononada gəlayniye, ammā av peštəpure ki toləša zəvon hammaysa rečina zəvone va hatto ingilisisa ğadime. Čəmə moa hamkoronan čoka miallimonin, ammā boməro čəmə moa hammaysa čoka iyan āğəlmānda miallime. Av hammay ženonsa xaymānd iyan rečine.

Anbarāni

Vist sorəsa vəyi ba ki čəmān nana mamlakatə lāp yola dānəšgāhanda koy kana. Av ingilisi muallime o əm zəvunə xayli čoke zəna. Nana hamiša baštān dars-ahāndun vonay ki ingilisi ila ča lāp ğania zəvununoy, vali av mutmaenne ki toləši hamasa māhbuba zəvune o de ingilisisan kānay. Čəmān nana hamkārunan čoka muallimin, vali māro čəmān nana gərdəsa čok o āğəlmānde. Av hama ženunsa mehrbān o māhbube.

English translation

My mother has been working in the country's biggest university for more than 20 years. She is an English language teacher and knows that language very well. Mother always tells her students that English is one of the richest languages in the world, but she's sure that the Talyshi language is the most beautiful language and is even older than English. My mother's colleagues are good teachers as well, but for me my mother is the best and smartest teacher. She's the kindest and most beautiful woman.

Footnotes

In memory of Zinaida Tabakyan

References

1 In my transcription of Northern Talyshi here, I almost entirely follow the transliteration system used by Wolfgang Schulze in his work Northern Talysh (Schulze, W., Northern Talysh (Muenchen, 2000)Google Scholar). It is worth noting that the symbols a and ā differ in quality rather than in length and denote front and back open vowels, respectively. The NT ə is a mid-high central vowel that should not be confused with the front a in Azerbaijani (and NT texts written in the Latin Azerbaijani alphabet) denoted by the same symbol. The other letters used here have their usual values (č = ç, ğ = ğ, x = x, š = ş, ž = j). The only difference between the system employed here and Schulze's work is that I use j as the Latin Azerbaijani c [dʒ], whereas Schulze represents it as ǯ.

2 Asatrian, G. and Borjian, H., ‘Talish and the Talishis (the state of research)’, Iran and the Caucasus 9.1 (2005), p. 43Google Scholar.

3 Clifton, J. et al., Sociolinguistic Situation of the Talysh in Azerbaijan (SIL International, 2005), p. 3Google Scholar.

4 See Bazin, M., Le Tâlech: Une Region Ethnique au Nord de l'Iran, tome 2 (Paris, 1980)Google Scholar; Stilo, D., ‘The Tati language group in the sociolinguistic context of Northwestern Iran and Transcaucasia’, Iranian Studies 14.3–4 (1981)CrossRefGoogle Scholar; Yarshater, E., ‘The Taleshi of Asālem’, Studia Iranica 25.1 (1996)CrossRefGoogle Scholar.

5 D. Paul, ‘A comparative dialectal description of Iranian Taleshi’ (unpublished PhD dissertation, University of Manchester, 2011), p. 18.

6 Pireyko, L., ‘Talyšskij jazyk. Dialekty Tati Irana’, in Osnovy iranskogo jazykoznanija: Novoiranskie jazyki: severo-zapadnaja gruppa, (eds.) Vasily Abaev, Mikhail Bogolyubov, and Vera Rastorgueva (Moscow, 1991), p. 91Google Scholar.

7 See Pireyko, L., Talyšsko-russkij slovar’ (Moscow, 1976), p. 351Google Scholar.

8 Bazin describes the population of Anbarān as bilingual, gradually becoming trilingual: ‘Nowadays, all this Ṭāleš population is bilingual, speaking both ṭāleši and Turkish, and increasingly becoming trilingual by using Persian as well’ (M. Bazin, ‘ANBARĀN’, Encyclopædia Iranica, online edition, 2012, available at http://www.iranicaonline.org/articles/anbaran (accessed 24 October 2023)).

9 Stassen, L., Comparative and Universal Grammar (Oxford, 1985), p. 24Google Scholar.

10 Treis, Y., ‘Comparative construction: an introduction’, in On the Expression of Comparison: Contributions to the Typology of Comparative Constructions from Lesser-Known Languages, Special Issue of Linguistic Discovery 16.1 (2018), p. iiiGoogle Scholar.

11 Dixon, R. M. W., Basic Linguistic Theory, vol. 3: Further Grammatical Topics (Oxford, 2012), p. 344Google Scholar.

12 For equatives and similatives, adopting the terminology used by Haspelmath et al., mark is replaced by standard-marker.

13 The Anbarān region consists of two rural districts (dehestān): the dehestān of Anbarān, which includes the villages of Anbarān-e Olyā, Qeshlāq-e Pelāzir, and Qeshlāq-e Sarābād, and the dehestān of Minābād, consisting of the villages of Jeyd, Kalash, Mirzānaq, and Minābād. The Anbarān region has a town under the same name, which was created in 1998 by uniting the villages of Anbarān-e Soflā and Aminjān (see Bazin, ‘ANBARĀN’). It should be noted that under the term ‘Anbarāni’, the Northern Talyshi dialect spoken throughout the Anbarān region is considered; however, this dialect itself includes a number of subdialects, which have some slight differences in terms of both phonology and grammar. This article is mainly based on the dialect of the former village of Aminjān, but the subdialects of other villages in the region have also been studied.

14 Stilo, ‘Tati language group’, pp. 138–41.

15 D. Stilo, ‘Two sets of mobile verbal person agreement markers in the Northern Talyshi language’, in Aspects of Iranian Linguistics, (eds.) Simin Karimi, Donald Stilo, and Vida Samiian (Newcastle, 2008), p. 364.

16 Pireyko, Talyšsko-russkij slovar’, p. 25, Pireyko, ‘Talyšskij jazyk’, p. 127; G. Windfuhr, ‘Cases’, in Encyclopædia Iranica, vol. 5 (Costa Mesa, 1992), p. 29.

17 B. Miller, Talyšskij jazyk (Moscow, 1953), p. 88.

18 Schulze, Northern Talysh, p. 17.

19 B. Miller (Talyšskij jazyk, p. 79) claims that the oblique case ending derives from Old Persian demonstrative particle hya. According to W. Schulze (Northern Talysh, p. 17), it corresponds to Old Persian -ahyā (genitive singular of a-stems), which in turn derives from Old Iranian case ending *-ahya (Windfuhr, ‘Cases’, p. 29).

20 In AnbTal, the postpositions ku and sa are used with personal pronouns and can attach either to the oblique forms (for 1SG and 3SG, e.g. mānəsa ‘on me’, avəku ‘from him/her’) or to direct forms (amaku ‘from us’, avonsa ‘on them’). In AzTal, however, these postpositions do not attach to personal pronouns, but instead attach to possessive pronouns (e.g. čəməku ‘from me’, əštəsa ‘on you’, čavonku ‘from them’, čaysa ‘on him/her’, etc.).

21 D. Paul, following Dixon, calls them ‘case clitics’ (for more details, see Paul, ‘Comparative dialectal description’, p. 158).

22 B. Miller suggests that the second singular and plural forms of possessive pronouns are also fused forms of čə and personal pronouns (əštə < čə + , šəma < čə + šəma). In the case of the second plural, the articulatory proximity of š and č has caused the latter to become inaudible (Miller, Talyšskij jazyk, p. 115).

23 For detailed information on the morphology of the verb in Northern Talyshi and especially the distribution of stem forms over the verb paradigm, see S. Kaye, ‘Morphomic stems in the Northern Talyshi verb: diachrony and synchrony’, in The Boundaries of Pure Morphology, (eds.) Silvio Cruschina, Martin Maiden, and John Charles Smith (Oxford, 2013), pp. 181–208.

24 See Stilo, ‘Two sets’.

25 See Paul, ‘Comparative dialectal description’, chapter 4.

26 Schulze, Northern Talysh, p. 98.

27 For information on the ergativity in Iranian languages in general and the course of its historical changes, see L. Pireyko, Osnovnye voprosy ergativnosti na material indoiranskyx jazykov (Moscow, 1968); T. Jügel, Die Entwicklung Der Ergativkonstruktion Im Alt- Und Mitteliranischen: Eine Korpusbasierte Untersuchung Zu Kasus, Kongruenz Und Satzbau: 21 (Iranica) (Wiesbaden, 2015); B. Scheucher, ‘Ergativity in new West Iranian’, in Essays on Typology of Iranian Languages, (eds.) Alireza Korangy and Behrooz Mahmoodi-Bakhtiari (Berlin/Boston, 2019).

28 R. M. W. Dixon, Ergativity (Cambridge, 1994), p. 1.

29 See Scheucher, ‘Ergativity in New West Iranian’, p. 6.

30 The NT enclitic pronouns are: 1SG. -əm, 2SG. -ə (AzTal)/-e/-ə (AnbTal), 3SG. -əš, 1PL. -əmon (AzTal)/-əmun (AnbTal), 2PL. -on (AzTal)/-en (AnbTal), 3PL. -əšon (AzTal)/-əšun (AnbTal). For more details about the functions of the enclitic pronouns, see H. Avchyan, ‘Enclitic pronouns in the Anbarāni dialect of Talyshi’, Bulletin of the Institute of Oriental Studies, 2.1 (2022), pp. 129–35.

31 When the adjective ends with a vowel, the attributive marker is usually omitted, e.g. gada ka ‘small house’, siyu mašu (AnbTal)/siyo mašo (AzTal) ‘black shoe’, etc. Adjectives ending with -i normally take the suffix -a, e.g. gadalia zua (AnbTal)/gadalia zoa (AzTal) ‘little boy’, ruania gužd ‘fatty meat’, etc.

32 The postposition sa ‘on, on top of’ has a nominal origin; it derives from the noun sa ‘head’ (< OIr. *sarah-). In both AnbTal and AzTal, besides its role as a postposition, it is used as a noun.

33 According to B. Miller, the postposition ku ‘from’ has a nominal origin from the noun ku ‘place’ (Miller, Talyšskij jazyk, p. 78). Daniel Paul states that in AzTal, ku ‘has word-level status as a nominal meaning “place,” in addition to its role as a postposition’ (Paul, ‘Comparative dialectal description’, p. 158), but I have never come across such a usage of this word nor do any of the main dictionaries of Northern Talyshi define ku as ‘place’ (F. Aboszoda, English-Talishi Dictionary (Muenchen, 2012); Pireyko, Talyšsko-russkij slovar’; F. Aboszoda, Russko-talyšskij slovar’ (Muenchen, 2012); B. Miller, Talyšskie teksty (Moscow, 1930)).

34 The preposition čə goes back to OIr. *hačā- ‘from’ (> New Persian az) (V. Rastorgueva and D. Edelman, Ėtimologičeskij slovar’ iranskix jazykov (Moscow, 2007), p. 301). This preposition is frequently used in AzTal, whereas its usage in AnbTal is sporadic. In fact, it is not even mentioned in the works dedicated to Iranian Talyshi, including Anbarāni (Paul, ‘Comparative dialectal description’; R. Amiriān-Budālālu, Guyeš-e Tāleši-ye Anbarān-e Ardabil [Talyshi dialect of Anbarān of Ardabil] (Hashtpar, 2005).

35 The postposition ada (AzTal) derives from OIr. *antar ‘in, among’ (> New Persian dar). The corresponding form in AnbTal is anda ‘in, with’, but it is not used in the constructions of comparison.

36 Miller, Talyšskij jazyk, p. 73.

37 In Anbarāni, the Persian comparative suffix -tar is very rarely used as the index of the normal comparative construction, when the standard and the mark are present. In my text corpus, it is found only once in the following sentence:

38 In the Persian language, when two items are being compared, the suffix -tar is affixed to an adjective in the predicate position, and the word az ‘from’ follows the item being compared and precedes the standard of comparison (see Shahrzad Mahootian, Persian (London/New York, 1997), p. 260). Examples (17), (18), and (19) translated into Persian would respectively be:

39 The verb pie ‘to love’ and ‘to want’ holds a special place in the verbal system of Northern Talyshi, showing an ergative pattern throughout its paradigm (see H. Avchyan, ‘A short story of Mullah Nasreddin in the Anbarāni dialect of the Talyshi language: text, translation, glossary and comments’, Bulletin of the Institute of Oriental Studies, 1.2 (2021), pp. 220–21).

40 In fact, this sentence is ambiguous. It can be translated either as ‘I love Maryam more than Žale (loves Maryam)’ or ‘I love Maryam more than (I love) Žale’. In some languages, such ambiguity arises because comparatives allow ellipsis of a repeated element, and it can be eliminated by restoring the missing elements, e.g. in English, the ambiguous sentence ‘I love you more than Ana’ can be paraphrased as ‘I love you more than Ana loves you’ or ‘I love you more than I love Ana’ (see Dixon, Basic Linguistic Theory, p. 368). In Northern Talyshi, paraphrasing this sentence is impossible, thus the meaning of the sentence has to be understood from the context.

41 Treis, ‘Comparative construction’, pp. vi–vii.

42 Dixon, Basic Linguistic Theory, pp. 343–75.

43 Ibid., p. 366.

44 Stassen, Comparative and Universal Grammar, pp. 24–52.

45 Ibid., p. 27.

46 Ibid., p. 41.

47 Treis, ‘Comparative construction’, p. x.

48 This sentence literally says ‘Tehran is a big[ger] city than all [cities] of Iran’. In this type of superlative construction, the words indicating a place can either be in the oblique case or marked by the postposition ada/anda ‘in’ (as in example (35)).

49 The only example in which the standard in the superlative construction is marked by the postposition ada is given by Pireyko (Talyšsko-russkij slovar’, p. 326):

50 Treis, ‘Comparative construction’, p. xii.

51 In Azerbaijani, the superlative is formed by placing the superlative word an ‘most’ before the adjectives. Examples (42) and (43) translated into Azerbaijani would respectively be (Latin Azerbaijani alphabet used):

52 Miller describes only one way to express the superlative in Northern Talyshi—the intensifier superlative discussed here further: ‘Only the descriptive method is used to express the superlative degree, the highest degree of any feature or quality – that is to put the adverbs ve “very” or xayli “a lot” before the adjective’ (Miller, Talyšskij jazyk, p. 73). He does not, however, provide any examples.

53 Pireyko mentions two ways for expressing the superlatives; the first is the intensifier superlative as described by Miller and the second is the absolute comparison superlative with the pronoun hamma/hammay ‘all, everybody’ (Pireyko, ‘Talyšskij jazyk’, p. 131). Meanwhile, Schulze does not discuss the superlative (Schulze, Northern Talysh, p. 19).

54 In the Persian language, when contrasting more than two different items, or the same item from more than two perspectives, the suffix -tarin is used. Superlative adjectives are attributive and precede the noun (Mahootian, Persian, pp. 260–61). Examples (44) and (45) translated into Persian would respectively be:

55 M. Gorshenin, ‘The crosslinguistics of the superlative’, in Neues aus der Bremer Linguistikwerkstatt: Aktuelle Themen und Projekte, vol. 31, (ed.) Cornelia Stroh (Bochum, 2012), pp. 65–88.

56 Treis, ‘Comparative construction’, p. xi.

57 Ibid., p. xii.

58 Ibid.

59 See Haspelmath, M. and Buchholz, O., ‘Equative and similative constructions in the languages of Europe’, in Adverbial Constructions in the Languages of Europe, (ed.) van der Auwera, Johan (Berlin, 1998), pp. 298301Google Scholar.

60 This word of Arabic origin also exists in Persian as qadr but, clearly, in both AnbTal and AzTal, it is an indirect Arabic loanword with a Turkic source as intermediary, cf. Az. qədər, Tr. kadar ‘extent, quantity’, also functioning as ‘as, as much as’.

61 Rett, J., ‘Similatives and the argument structure of verbs’, Nat Lang Linguist Theory, 31.November (2013), p. 1125CrossRefGoogle Scholar.

62 Haspelmath and Buchholz, ‘Equative and similative constructions’, p. 309.

63 Ibid., p. 290.

64 Haspelmath, M. et al., ‘Equative constructions in world-wide perspective: a crosslinguistic perspective’, in Similative and Equative Constructions: A Cross-linguistic Perspective, (eds.) Treis, Yvonne and Vanhove, Martine (Amsterdam/Philadelphia, 2017), pp. 1415Google Scholar.

65 Haspelmath and Buchholz, ‘Equative and similative constructions’, p. 278.

66 Avchyan, H. and Voskanian, V., Tališereni dasagirk (Yerevan, 2022)Google Scholar.

Figure 0

(1)

Figure 1

(3)

Figure 2

(4)

Figure 3

(6)

Figure 4

(8)

Figure 5

(9)

Figure 6

(11)

Figure 7

(13)

Figure 8

(15)

Figure 9

(16)

Figure 10

(17)

Figure 11

(20)

Figure 12

(22)

Figure 13

(24)

Figure 14

(25)

Figure 15

(26)

Figure 16

(29)

Figure 17

(31)

Figure 18

(32)

Figure 19

(36)

Figure 20

(38)

Figure 21

(42)

Figure 22

(44)

Figure 23

(46)

Figure 24

(52)

Figure 25

(54)

Figure 26

(57)

Figure 27

(59)

Figure 28

(62)

Figure 29

(66)

Figure 30

(68)