Ðàçìåùåíî íà http://www.allbest.ru/

Towards a typology of close appositional constructions with proper names



Abstract

proper names typological language

The present work is devoted to the syntax of constructions of specification with proper names (i.e. close appositional constructions in English) in a typological perspective. The provided typology is based on the results obtained in the analysis of the existing grammar descriptions and available corpus data in 95 languages. The paper discusses the morphosyntactic means that languages use to express specification, namely, juxtaposition, attribution, and other less common strategies. The separate sections deal with the distribution of case marking and the order of components in appositive phrases. The work includes an independent study in the spirit of intragenic typology -- a corpus-based analysis of factors influencing variation in case agreement of appositives in six Slavic languages (Ukrainian, Belarussian, Polish, Czech, Croatian, and Slovenian). The work presents evidence that appositions are generally treated as one NP in the languages of the world, although there are facts that shows that appositions possibly constitute syntactic units both “smaller” and “bigger” than NP. Evidence from the languages using clearly subordinative (“attributive”) strategy in constructions of specification shows that a common noun is generally considered to be the syntactic head of a construction, but there are certain exclusions.

Key words: close apposition, proper names, linguistic typology, corpus linguistics, case marking, word order, constituency



Table of contents

Basic principles of information delivery

List of abbreviations

Morphological glosses

1. Introduction of close apposition with proper names

1.1 Towards a definition of nominal close apposition and introduction of the notion construction of specification

1.2 On the previous approaches to close apposition

1.3 Aims and objectives of the study

1.4 Limitations

1.5 The outline

2. Methodology

2.1 The sample

2.2 Methodology and sources

3. Expression of specification in the languages of the sample

3.1 Juxtapositional strategy and its syntactic properties

3.2 Attributive strategy

3.2.1 Attributivization parallel to possessive constructions

3.2.2 Attributive strategy different from that found in possessive constructions

3.3 Other strategies to express the meaning of specification

3.3.1 Constructions of specification parallel to relative clauses

3.3.2 A special marker of apposition

3.3.3 Representative particle

3.3.4 Unclear cases

3.4 Conclusions

4. Case marking in appositive constructions in typological perspective. The problem of constituency

4.0 Why interaction of apposition with case marking is worth considering

4.1 Patterns of case marking in appositive constructions of specification and constituency

4.2 Towards a better understanding of constituency

4.3 Case marking in Slavic appositive constructions

4.4 Conclusions

5. Order of constituents in constructions of specification in a typological perspective

5.1 Conclusions

6. A sketch on definiteness in appositive constructions

References

Appendix 1

Appendix 2



Basic principles of information delivery

In the present work, all examples, except for examples in English, contain four types of lines: (0) the line giving information on the genetic affiliation of a language as well as the source of the given example; (1a) the line with an example in a spelling adopted in the original publication from which the present material is drawn; (1b) the string with a transcription in Latin (given only in the case the original script is impossible to read without a special knowledge, like Japanese or Georgian); (2) morphemic glosses, (3) translation. When quoting examples from grammars, the morphological glosses proposed in these works remain unchanged. That means that sometimes functionally identical entities are glossed differently in the present work depending on the glosses used in the original publications. Conversely, different markers are sometimes glossed identically due to the same reasons. The actual functions of kind the morphemes should be clarified in the source works. The list of the glosses used is given below. When quoting examples from publications that are not in English, the glosses are translated into English without preserving the original text. In the absence of glosses in the cited source, morphological analysis (wherever possible) is carried out by me in consultation with the available grammatical descriptions. Whenever any romanization of the Cyrillic script is present, it is made in accordance with the international standard ISO 9 https://www.iso.org/standard/3589.html.

List of abbreviations

EPG -- Ereduzko Prosa Gaur

RNC -- Russian National Corpus

WALS -- World Atlas of Linguistic Structures

SE -- Sketch Engine



Morphological glosses

1, 2, 3	first, second, third person
a	absential
abl	ablative
abs	absolutive
acc	accusative
add	adessive
adi	additive
all	allative
ao	actor orientation
aor	aorist
appl	applicative
apud	apudessive
art	article
asc/ assoc Existence of more than one variant of gloss for the same grammatical meaning is due to different notational conventions in different grammars cited.	associative
ascm	associative modifier
atten	attenuative
atr/a/ att aux	attributivizer/ attributive form auxiliary verb
ben	indirective object
ca	connective adverb
cl	classifier
clt	clitic
cng	connegative
cm	compact/ class marker
cmp	completive
comp conj	complementizer coordinating conjunction
core	non-nominative core article
cs	construct state
dcl	declarative
dat	dative
del	delative
dem	denomstrative
def	definite article
des	destantive
el	elative
erg	ergative
ex	exclusive
f/f	female
fl Occurs with flexible items (paper, cloth, etc.), but is also used for items that are out of sight, and also as a kind of default article (Marlett, Moser 2004: 24).	flexible, out of sight, or default
foc	focus
gen	genitive
gpf	gap filler
goal	goal
habit	habitual
hes hon	hesitation honorative
imm	immediate
inf	infinitive
intr	intransitive
io	indirect object
ip	intransitive verb prefix
ipsr	?rst person inclusive possessor
itr	intransitive copula
iter	iterative
izaf	izafet
iter	`iter' (from lat. `way')
lnk	linker
m	male
me	male ego (relating to kinship terms)
mid	middle affix
mod	modified
mot_purp	motion purpose
n1	non-first person
nd	near distal deictic
nondecl	non-declarative pronoun
nom/ nomin neg	nominative negation
obl	oblique
p/poss/pssr pc/prt/ptcp	possession participle
perl	perlative
pf/ pft	perfect
pl/ plur	plural number
pln	place name
pluhum	human plural
post	postposition
prl	perlative
pres	presentative particle
prg/prog	progressive
pro	pronoun
prol	prolative
pv	passive
prs/pres	present tense
psn	person name
pst/a	past tense
red	reduplication
rel	relativizer/ relator
rem	remote past
rep	reportative
s	singular
sim	simultaneous
sn	subject nominalizer
spir	same subject + irrealis
srdir	superdirective
stats	subject of a stative verb
subj	subject
sw	switched reference
xpsr	?rst person exclusive possessor
top	topic
vnnm	nominalizer
w	witnessed



1. Introduction of close apposition with proper names

1.1 Towards a definition of nominal close apposition and introduction of the notion construction of specification

The present work is devoted to the study of the syntactic organization of nominal close appositional constructions with proper names (also called restrictive appositional constructions) with the methods and tasks of linguistic typology. Under the term of close apposition, the following constructions in English (1) and Russian (2) are generally understood:

(1) Now, let me praise the American writer Dickey. (Thesing, Wrede 2009: 13)

(2) ... film <...> gde zvuèali pesni na stixi poet-a Vožnesensk-ogo

film where sound:pst;pl song:pl on poem:pl poet-gen Voznesensky-gen

`a film <...> where there were songs by poet Voznesensky'.(Razzakov 2009)

In the examples (1) and (2), taken from English and Russian contemporary prose, the boldfaced nominal phrases represent close appositions formed by juxtaposition of two nouns (writer and Dickey in (1) and poeta and Vožnesenskogo in (2)). In the existing works on close apposition, constructions of this type are most commonly defined as consisting of two nominal constituents (one of which is most commonly a proper name), which (i) have a common referent (a particular person called Dickey and working as a writer) and (ii) take apparently the same syntactic position in a sentence (a direct object position in (1) and (2)). The semantic (referential) and syntactic parallelism between parts of a construction are thought to be the key features of this special type of grammatical relation as opposed to coordination and subordination (Acuña-Fariña 1999: 454).

Close appositional constructions are commonly analyzed in comparison with the so-called loose appositional constructions (Herringa 2012). Contrasting the two types of apposition allows to name at least the following most basic properties of close appositions:

  Parts of a close appositional construction constitute one intonation group, while loose apposition is characterized by a marked violation of the intonational integrity of the phrase: the parts are separated by a pause or change in a tonal counter or pitch (Dehe 2014: 55-63) The prosodic break is not the only intonational property that allows to distinguish between close and loose apposition in English. In close apposition the stress pattern is said to be “secondary-primary”, which means that the primary stress is placed on the proper name whereas the preceding common noun only receives secondary stress (Haugen 1953; Francis 1958). In loose apposition, both nominal elements receive primary stress (Keizer 2007: 26).. This can be seen in the following pair of examples:

(3) a. My cousin Bill now lives in Boston. < close apposition

b. My cousin, Bill, now lives in Boston. < loose apposition

Although the intonation is often the only formal (at least, noncontroversial) property that allows one to distinguish between close and loose apposition, the contrast found at the level of prosody is considered to be a superficial correlate of the semantic difference between the two types of constructions.

  Close appositions, in the general case, have the so-called restrictive (limiting) semantics, while loose appositions have non-restrictive (specifying) semantics.

In other words, in restrictive (close) appositions, the reference that the structure as a whole has is more specified than the reference of each of its parts taken separately. Thus, many authors have argued (Heringa 2012: 3; see also Quirk et al. 1985: 1303-1304, among others) that in (3)b a speaker has only one cousin. The fact that his name is Bill constitutes a supplementary information added by a speaker for some reason (may be when the addressee knows Bill by name and the speaker tries to ensure that the addressee knows who is talking about). In (3)a there can be more cousins, but the one meant is called Bill. Although the distinction is understandable based on pairs like the one in (3), and widely discussed, it is far from being absolute due to the existence of examples like my only brother Barry (SE), where both parts of the close appositional structure my only brother and Barry are supposed to have unambiguous reference.

Appositional constructions do not represent a homogeneous group. Acuña-Fariña in (Acuña-Fariña 1999; 2006; 2009; 2016) develops a theory for apposition regarding it as a family of patterns that need to be analyzed separately because of the formal syntactic dissimilarities and different semantic relationships between the elements. He postulates 10 types of nominal appositive structures in English (Acuña-Fariña 2009: 455). However, since each language supposedly gives its own reasons to distinguish one type from the other, it is more feasible to cluster Acuña-Fariña's types in the following more general groups:

  appositions with “true” proper names as a second (or less commonly, first) element: the king George, the river Thames, gorod Moskva `the city of Moscow'(Russian));

  appositions made up of two common nouns (my friend the poet);

  appositions with personal pronouns (we boys).

Furthermore, the focus of the present work is on appositive constructions consisting of a common noun, serving as a generic term, or a sortal as it is defined in (Matushansky 2013)), and a proper name, specifying the reference. Loose appositive constructions, as well as other varieties of close apposition, are brought into discussion if necessary.

In what follows I will not use the term appositional construction in the sense it is most often used in the existing studies on the subject and as it was mainly introduced above for the following reasons. First, the term apposition, in addition to its usage for the description of English binominal noun phrases (NPs), as in the examples above (1), which is the narrow meaning of the term, is often applied to any sequence of two constituents concatenated without any linking element. In this case only syntactic form is determined, whereas semantics can vary. Among the most widely cited semantic relations expressed through apposition in its syntactic sense is possession (which is mentioned in many papers, among which (Dryer 2006:31)) and a number of other relations between two nominals inside of an NP which were claimed to be included in the set of prototypical possessive relations in (Aikhenvald, Dixon 2013: 3) (kinship, part-whole, material, quantity). When used in this sense, apposition is synonymous to the term juxtaposition. Second, as will be shown below, apposition, in its syntactic sense explained in the previous passage, is not the only possible way to express specification. In order to avoid the terminological ambiguity and confusion I will introduce the comparative concept of construction of specification with the following key properties:

  it consists of at least two nominal components, one of which is a proper name To give a typological definition of proper name is not an easy task (Langendonck, de Velde 2016). In the present work I am guided by the intuition of the authors of the original grammar descriptions and mainly rely on the orthographical criterion (capitalization)., and the other is a common noun;

  there is a relation of specification between the components in the way that the common noun introduces a class of referents (writers, grandmothers, cities, lakes, etc.) and the proper name specifies the reference by pointing out a particular individual among the class of items indicated by the more general nominal (Hemingway, Elizabeth, Moscow, Baikal).

The term appositional/ appositive construction (CA) is used further to refer to only those constructions of specification that are syntactically built by simple juxtaposition of their constituents. The term juxtaposition is used in the broadest sense in relation to any sequence of nominal elements that appear to be concatenated without any morphosyntactic means of subordination or coordination. However, since the present work is mainly inspired by the problems that were raised while studying close appositions in English and other closely related languages, the range of objectives of this study clearly addresses the questions most relevant to appositions in their syntactic sense, namely: constituency, headness and order of constituents. The term appositive(s) is used to refer to the components of an appositive phrase separately.



1.2 On the previous approaches to close apposition

Before outlining the aims of the present research, it is important to consider the range of questions generally discussed in relation to apposition in the existing literature, for the reason that they to a large extent determine the nature of the goals that are pursue in the further work.

In addressing the question of syntactic nature of close appositional constructions various analyses have been proposed, with the main attention to the data of English. As has been already said in the previous section, authors generally agree that close appostitions (i) involve the juxtaposition of at least two nominal elements, one of which is, most commonly a proper name, (ii) constitute one intonation unit and (iii) have restrictive semantics (as opposed to loose apposition). At the same time, there is an ongoing discussion concerning both the range of structures to be regarded as genuine close appositional constructions, and the particular issues of morpho-syntactic properties of CA, such as constituency, headness, referentiality and (less commonly) case agreement.

In the pioneering studies on close apposition (Potsuma 1904: 311-320; Lee 1952: 268-275) attention was mainly drawn to the prototypical “the poet Burns” type of CA, regarded as a representative of CA at large (although some other examples like the star Sirius, the drug aspirin were considered as well). Close appositions were clearly distinguished from what is called `quasi adjectives' (the tyrant Caligula) and dvandva expressions (the soldier king). Not surprisingly, syntactic headness (the syntactic primacy of one element in the group over the others) was seen as the main theoretical challenge represented by the structures of this type. At first, the internal dependency in this construction was mainly considered to be a subtype of modification with the first noun (defined positionally = common noun) being the head and the second noun (defined positionally = proper name) serving as modifier. The primacy of the proper name was justified by the semantic consideration that the second noun limits or defines the meaning and the reference of the first. The same approach is also applied in some papers on apposition in Russian (Apresjan 2010: 34). There are also salient morpho-syntactic evidences proving that the common noun is the morphosyntactic locus of a CA construction (Zwicky 1985): (i) pluralization (the biologist-s_pl Huxley (Keizer 2007: 56)) and subject-verb agreement (the biologist-s_pl Huxley were_pl (ibidem: 58))

On the contrary, Haugen (Haugen 1953: 165-170), although following the theory of modification, argued that it is the second noun which is the head (“the thing being talked about”) in English CAs. He based his argument on the semantic primacy of the proper name, as well as on the observed stress pattern secondary-primary, though the last assumption is based on a subjective intuition and has not been justified by any empirical data till now. The semantic-syntactic evidence considered by Haugen was the possibility to omit the common noun without any change in meaning (the poet Burns is the same referent as the one introduced by Burns, but not by the poet).

A third perspective is given by Hocket (Hocket 1955: 99-102), who suggests an exocentric (no distinguished head -- modifier relation) structure for close appositions with the two elements semantically serving as “modifiers to each other”. Although Hocket's approach was not fully theoretically elaborated and poorly grounded, it is reflected in the comprehensive account of (Quirk et al. 1985), where the apposition is described as the relation “resembling coordination” of elements of the same rank.

Burton-Roberts' (Burton-Roberts 1975) account was the first analysis of apposition within the transformational framework. He considers appositions to be derived from an underlying construction of the type [[Burns][Burns is the poet]] The full transformational path is described in the original work (Burton-Roberts 1975)., where the poet is the complement of a copular verb and therefore cannot be referential and cannot serve as the head. Consequently, the relation of the members of apposition is modifier (common noun) -- modified (proper name).

In a comprehensive corpus-based study by Meyer (Meyer 1992: 10-54) the range of constructions described as appositional was considerably expanded on the basis of real examples observed in the corpus. According to Meyer, different types of appositions form a syntactic as well as semantic scale from the most prototypical examples (the poet Burns type) to the more peripheral ones (Burns the poet; we the poets), deviating from the core in the range of possible syntactic transformations (particularly, inversion) and their meaning. For Meyer, it is impossible to suggest a uniform constituent structure for all the different types of constructions, therefore his main thesis was that apposition is an exceptional kind of syntactic relation intermediate between coordination and attribution.

This not `all-purpose' approach of Meyer was later elaborated in the more recent accounts on apposition as well. The “early” Acuña-Fariña (1996) As I have no access to the work byAcuña-Fariña of 1996, I rely on the summary given in (Kojadinovicì 2018). states that the expressions described as close appositives in the literature cannot be analyzed uniformly and should rather be divided into several types each representing its own semantic and syntactic properties. Acuña-Fariña's position on the constituency and headness in phrases like the poet Burns is that appositions of this sort generally constitute one NP, and the relation between the components is modifier (common) noun -- head (proper name). He establishes that as a proper name is inherently definite, it imposes restrictions on the class of modifiers that can cooccur with it and with close appositions as well (cf. the poet Burns/ *a poet Burns and the famous poet Burns/ *the tall poet Burns as the famous Burns/ ^?the tall Burns) (via (Kojadinovicì 2018)).

In the later works (Acuña-Fariña 2009, 2016), Acuña-Fariña develops his theory of close apposition within the construction grammar framework, representing different types of appositives as a family of constructions. What is different about his later account is that he diligently summarizes the arguments that had been proposed in the literature for the headness of either the first element (pluralization the artist*(s) Van Gogh(s) and agreement) or the second (phrasal stress and selection restrictions) and argues that there is no clear evidence which of the elements is the head, so that the pattern of close apposition is simply “unresolved” (Acuña-Fariña 2009: 468).

Keizer's (Keizer 2005) paper attempts to revisit the formal analysis of close apposition with the focus on its usage in discourse. Keizer describes four discourse functions that appositional constructions can have. According to her, the use of one or another particular syntactic structure of CA can be to a large extent determined by its usage conditions, such as its function introducing a new participant or otherwise grounding a referent to the preceding discourse. An important point of Keizer's analysis was that close appositions can be preceded by indefinite articles in English, as their core function is introduction, and referents in need for introduction are expected to be new and indefinite. At the same time, such examples are extremely infrequent in the corpora, which explains the fact that this possibility wasn't even considered in previous works. In (Keizer 2007), it is explicitly argued, in contrast to the existing papers, that nouns in apposition are not referential on their own and cannot form NPs independently, but rather constitute a single NP together: *[NP] [NP_P] vs. det [[N] [N_P N_p stands for Proper Name in this case.]] (ibidem: 38).

From the brief overview above it can be seen that the central questions addressed in relation to apposition in the literature are (i) headness (which of the elements is the syntactic head), (ii) constituency (does apposition constitute one NP, or two distinct NPs or neither) and (iii) definiteness in appositive phrases. The main problem of the existing discussion appears to be that it is very limited in terms of the actual linguistic data that is considered. One of the main tasks of the present work is to contribute to the debate by introducing the data from languages other than English and addressing these issues on the basis of new and more revealing data.



1.3 Aims and objectives of the study

The aims of the present study are the following. First, to examine the inventory of morphosyntactic means used in the languages of the world to express the relation of specification between two nominals, one of which is a common noun and the other is a proper name, and to produce a typology of the observed strategies. Second, on the basis of this data, to consider (i) if there are any typologically justified reasons to treat appositive structures as constituting one NP or separate NPs and (ii) if there is any evidence for one element having the structural priority (thus, being the head). To attain those objectives, it appears to be reasonable to not only examine the constructions parallel to English apposition in their structure and function, but also to consider the interaction of close apposition with other syntactic phenomena such as case marking. The more detailed introduction onto each of subtopics is given in respective parts of the present work.

A naturally arising question about the constructions of specification is why languages use them at all and whether some languages do that more often than others. The source of these questions is quite a reasonable belief that the constructions of this type are a phenomenon of redundancy, which is generally thought to be avoided in the language, since there is no functional need in majority of cases to name the same referent using more than one description. In most grammars having a separate paragraph on the constructions of specification (which is most often not the case ), authors merely mention that constructions of this sort allow to specify the reference of a common noun (see, for example, (Ingram 2001: 300-302) on Anamuxra, (Aikhenvald 2003: 480-481) on Tariana among others). Generally, the question of a discourse related use of constructions of specification is not addressed in grammars at all, so that it is hardly possible to give any typology of their discourse functions. Nevertheless, it can be said for sure that in some languages, apposition is functionally more required than in others. For example, Donohue (Donohue 1999) mentions about Tukang Besi (an Austronesian language in Indonesia) that a pronoun or a proper name cannot be possessively modified in that language. When this is required, an appositive construction is used instead:

(4) Tukang Besi (< Celebic < Austronesian, (Donohue 1999: 132):

a. te ia [te [Wa Si? i]_np [obu-su]_np]_kp

core 3sg core Wa Si ìi dog-1sg.poss

`That's Sii, my dog'.

b. *te ia [te [Wa Si? i[-su]_poss]_np]_kp

core 3sg core Wa Si ìi-1sg.poss

int. `That is my Si ìi'

Although the motivation for apposition is quite clear in examples like (4) and can potentially be universal, I have not come across any other mentions of the similar restrictions. I also struggle to say if any language avoids using constructions of specification at all. The number of grammars I looked through is far bigger than the number of languages mentioned in this work, and majority of grammars lacked any description of constructions of specification, but no one of them claimed that such constructions to be ungrammatical. Thus, the question of functions of apposition remains open till now and may require a special study.

1.4 Limitations

The scope of this study was limited in terms of non-exhaustive description of the constructions of specification in existing grammatical descriptions. The examples of the constructions under interest generally come from two types of sources. The first type is represented by descriptions of apposition (or other relevant grammatical phenomena) in a grammar of a particular language or an article dealing specifically with this type of expressions. The problem is that due to the lack of explicit indicators of grammatical relation between the parts of the construction, appositions have long been considered insufficiently “interesting” to be studied and were often ignored by researchers. That is why, relatively detailed sections on apposition are only found in the grammars of either those languages for which this type of relation in the NP is the only one possible (as is the case in many Australian languages (see Section 4)) or at least quite common beyond the constructions of specification (in this case, however, relevant examples on constructions with the specificationional semantics very often are missing), or in the grammars of well-described European languages. This fact hinders any in-depth analysis to a large extent. The second source of data is represented by single examples of constructions of specification with proper names found sporadically in grammars and generally not accompanied by any analysis or in the Bible translations. In this case the problem again is that such contexts are very rare to find, and when found, one cannot be sure if they represent the only (or even the most basic) way to express this meaning in the language. The scarcity of the available data resulted in the main shortcoming of the present work, which is the not homogeneous character of the contexts considered. That is, for different languages lexically different types of constructions were sometimes analysed. Looking for a representative of a construction of specification with names of humans, in some cases, I used examples including such common nouns as king, president, professor, writer etc. in the position of a common noun. These nouns describe profession or social status of a referent. In other cases, I used examples with kinship terms and quasi kinship terms, such as friend, brother, father etc. In the end, in cases when almost exclusively the data from the Bible translation was analysed, I used expressions such as prophet Isaiah, which is a special case of a religious title. As will be shown below, lexico-semantic properties of constituents can influence the parameters of syntactic organization of construction of specification. That is why, the non-homogeneous character of the data complicates the comparison to a large extent. However, it appears that there is no other possibility to gather a sufficient amount of data for a balanced typological sample of languages.



1.5 The outline

The present work consists of five sections (excluding the introductory section). Section 2 gives information on the sample of languages, on which the present study is based and discusses the general methodology. Section 3 gives an overview of strategies that are used languages of the world to express the meaning of specification. Section 4 addresses the question of distribution of the case feature in (syntactically) appositive constructions of specification in the languages with case. Section 4.5 is devoted to the discussion of case agreement in Slavic languages. In Section 5 the patterns of constituency ordering in appositive constructions are observed. Section 6 is a sketch of the problem of definiteness in the constructions of specification. Section 7 is the conclusion.

1.6 Acknowledgements

I would like to express my gratitude to my extraordinary supervisor Masha Kholodilova, who made linguistics possible for me. Being extremely inspiring, generous with her ideas, curious and strict, she at the same time was always very supportive and understanding. I would be very far from where I am now if it wasn't for her. All the weaknesses of the present work are on my conscience.



2. Methodology

2.1 The sample

This study of constructions of specification appears to be almost the first effort to approach a part of syntax of noun phrase that has never been profoundly investigated from a typological point of view (except for my own previous work and work by Zubova (Logvinova 2019a; Zubova 2018) that both mostly focused on the languages of Eurasia). That is why the sample of languages should aim to cover as much of linguistic diversity as possible. The sample used in this study is genealogically stratified at the level of genus (Dryer 1989). Genus is a group of languages with a common ancestor having a time-depth of about 3500-4000 years before present (Oskolskaya et al. 2019). The number of genera per family is beyond the control, but it generally follows the principle that larger families are represented by more genera. The genealogical classification used in WALS is taken as referential.

A very important note about this sample is that it includes most of the languages that I used in (Logvinova 2019a) and that come from the WALS 200-language sample The full list of the languages included can be consulted on the WALS web-page: https://wals.info/languoid/samples/200 (with a number of elaborations that are discussed in (ibidem)). There are exactly 41 such languages out of 95 in the present sample (they are given in italics in Table 1), and big part of them (30) are found in Eurasia. The sample used in (Logvinova 2019a) was both areally and genetically skewed, as most of the languages considered were from the Indo-European family (12). Moreover, for some genera, more than one representative language was included (for example, Estonian and Finnish from the Uralic stock, Turkish and Chuvash from the Turkic family). The sample used in the present work includes languages from the five remaining macro-areas (namely, Africa, North America, South America, Australia and Papunesia). When developing the final sample, as there was a special interest in the languages having case marking, the sample on the position of case markers in world languages (Dryer 2013) was used as a starting point. However, as the decision was taken to only include in the sample those languages for which it was possible to find information on how constructions of specification with personal names of humans as well as with toponyms are built, it was necessary many times to switch from the language introduced in the sample to a closely related one for which there were more sources. As a result, the final sample is not inherent to any ready-made sample in WALS or any other work. That, of course, leads to certain shortcomings. Although the current sample is more genetically and geographically equalized, its main shortcomings are the following:

§ The number of Indo-European languages is still disproportionally big in relation to other language families, and the sample as a whole is skewed towards the languages of Eurasia. According (to the Genus-Macroarea method introduced in (Miestamo et al. 2016), the expected number of Eurasian languages in a 100-language sample is 8, while the present sample includes 32.

§ The sample is bibliographically biased, as it includes languages for which more elaborated grammar descriptions were available and for which it was possible to find the required data. That means, that if there was no information on constructions of specification in the available descriptions and any other written sources could not be found and it was impossible to find native speakers of that language, the language was replaced by a closely-related one, for which more information on contexts of specification was available. This methodological downside is not critical, as it seems quite reasonable to assume that the absence of a special treatment of constructions of specification in a grammar is the consequence of expressions of this sort not representing any interest or challenge to the researcher in being very similar to the constructions of specification in English.

In Table 1 languages are organized both geographically (according to macroareas) and genealogically (according to language families and genera).



Table 1

Genealogical distribution in the languages of the sample The languages that are given in italic font were used in (Logvinova 2019)

	Macroarea	Linguistic family	Genus	Language
	Eurasia (32)
1		Indo-European	Germanic	German
2			Slavic	Slovene
3			Indic	Hindi
4			Romance	Romanian
5			Baltic	Latvian
6			Celtic	Irish
7			Albanian	Albanian
8			Armenian	Armenian
9			Greek	Greek
10			Iranian	Persian
11		Uralic	Finnic	Finnish
12			Ugric	Hungarian
13			Samoedic	Tundra Nenets
14		Altaic	Turkic	Chuvash
15			Tungusic	Evenki
16			Mongolic	Khalkha
17		Nakh-Daghestanian	Nakh	Ingush
18			Lak-Dargwa	Lak
19			Avar-Andic-Tsezic	Khwarshi
20			Lezgic	Lezgic
21		Dravidian	Southern Dravidian	Kannada
22		Sino-Tibetan	Bodic	Ladakhi
23			Kuki-Chin	Mizo
24		Northwest Caucasian	Northwest Caucasian	Abkhaz
25		Chukotko-Kamchatkan	Northern Chukotko-Kamchatkan	Chukchi
26		Kartavelian	Kartavelian	Georgian
27		Yeniseian	Yeniseian	Ket
28				Yukaghir (Kolyma)
29				Basque
30				Korean
31				Nivkh
32				Japanese
	Africa (22)
33		Afro-Asiatic	Semitic	Arabic (Ñlassic)
34			West Chadic	Hausa
35			Biu-Mandara	Gude
36			Southern Cushitic	Iraqw
37			North Omotic	Wolaytta
38		Central Sudanic	Moru-Ma'di	Lugbara
39			Bongo-Bagirmi	Bargimi
40			Nilotic	Teso
41			Nubian	Nubian (Dongolese)
42		Niger-Congo	Adamawa	Doyayo
43			Kwa	Lelemi
44			Bantoid	Swahili
45			Ubangi	Sango
46			Gur	Supyire
47			Northern Atlantic	Wolof
48			Defoid	Yoruba
49		Mande	Eastern Mande	Mano
50		Saharan	Western Saharan	Kanuri
51		Songhay	Songhay	Koyra Chiini
52		Berber	Berber	Berber (Siwa)
53		Aymaran	Aymaran	Aymara
54		Khoe-Kwadi	Khoe-Kwadi	Nama
	North-America (11)
55		Muskogean	Muskogean	Koasati
56		Mayan	Mayan	Huastec
57		Mixe-Zoque	Mixe-Zoque	Popoluca (Sierra)
58		Haida	Haida	Haida
59		Hokan	Seri	Seri
60		Iroquoian	Northern Iroquoian	Mohawk
61		Keresan	Keresan	Acoma
62		Penutian	Sahaptian	Northwest Sahaptin
63		Siouan	Core Siouan	Crow
64		Uto-Aztecan	Aztecan	Nahuatl Clasical
65		Oto-Manguean	Chinantecan	Chinantec
	South America (15)
66		Panoan	Panoan	Chácobo
67			Tacanan	Ese Ejja
68		Arawakan	Southern Maipuran	Yine
69			Northern Maipuran	Tariana
70			Central Arawakan	Paresi-Haliti
71		Chapacura-Wanham	Chapacura-Wanham	Wari'
72		Jivaroan	Jivaroan	Huambisa
73		Tupi-Guaraní	Tupari	Akuntsu
74			Tupian	Cocama-Cocamilla
75		Movima	Movima	Movima
76		Quechuan	Quechuan	Quechua (Huallaga)
77		Cariban	Cariban	Tiriyo
78		Urarina	Urarina	Urarina
79		Chapacura-Wanham	Chapacura-Wanham	Wari'
80		Choco	Choco	Epena
	Australia and New Genia (7)
81		Pama-Nyungan	Western Pama-Nyungan	Martuthunira
82			Central Pama-Nyungan	Arrernte (Mparntwe)
83			Northern Pama-Nyungan	Warrongo
84		Yangmanic	Yangmanic	Wardaman
85		Sepik Hill	Sepik	Alamblak
86		Trans-New Guinea	Binanderean	Suena
87			Madang	Anaamuxra
	Papunesia (8)
88		Torricell	Kombio-Arapesh	Arapesh (Mountain)
89		Gunwinyguan	Gunwinygic	Bininj Gun-Wok
90		East Bougainville	East Bougainville	Motuna
91		Yangmanic	Yangmanic	Wardaman
92		Dagan	Dagan	Daga
93		Austronesian	Northwest Sumatra-Barrier Islands	Gayo
94			Oceanic	Samoan
95			South Halmahera - West New Guinea	Taba

The geographical distribution of the languages of the sample is presented in Picture 1.



Picture 1 Geographical distribution of the languages in the sample

In some cases, I also discuss information on languages that are not included in the main sample as it was impossible to find all the necessary contexts for them, but the available data on them is worth discussing.

2.2 Methodology and sources

The methodology of this study can be described as follows:

1. For each language of the sample, I collected the necessary information on a list of macro parameters defining the main properties of NP architecture. The list of macro-parameters includes the following characteristics:

- basic word order in the clause;

- the order of the head noun and its genitival and adjectival dependents;

- the presence / absence of case;

- the presence / absence of case agreement of the adjective with the head noun in the NP.

The data for the first three parameters was collected based on the information provided in WALS. If the data for any of the parameters was missing, the information was additionally clarified in the available grammar description. In some cases, the data from WALS was adjusted according to the data in a trusted grammatical descriptions. The information on the last parameter (presence / absence of case agreement) was collected directly from grammars of the particular languages.

2. Furthermore, for each language, I (1) defined how the constructions of specifications with (i) proper names of humans and (ii) toponyms are built and (2) what is the basic order of constituents in constructions of specification of both types.

The data considered in the present study mostly comes from the following sources:

  available grammar descriptions;

  existing original texts of any kind;

  Bible translations (the list of contexts used when working with the Bible is given in the Appendix 3);

  monolingual and parallel corpora;

  consultations with native speakers and language experts.

When working with grammars, I first looked for the special sections or paragraphs dealing with apposition, and then (independent from the result obtained on the first stage) I searched for words that were the most expected to be used in constructions of specification of the needed types, i.e. the nouns that are most often accompanied by a proper name of a person (`king', `friend', `poet' etc.) or toponym (`city', `village', `town', `river', `mountain' etc.).

It is important to note on what types of constructions were not considered as constructions of specification. The initial restriction was connected to what is called `quasi-' constructions of specification in the present study. An example of a `quasi-' construction is given below:

(5) Seri (< Seri < Hokan, (Marlett, Moser 2004: 475)

heezitim caacoj [Jericoì hapaìn quij

city sn-big sn-pv-say the.cm

`the city of Jericho', lit. `the city called Jericho'

The structure in (5) has its parallels in English and Russian as well. But in these languages, this structure is not the main and far not the most frequent way to express specification. Consider an example from Russian:

(6) Russian (< Slavic < Indo-European, personal knowledge)

a. devoc?k-a Rita

girl-nom.sg Rita

`the girl Rita'

b. devoc?k-a po imen-i Rita

girl-nom.sg prol name-loc Rita

`a girl called Rita', lit. `a girl with the name Rita'

Althouh I came across such constructions several times, they were never claimed to be the only possible way to express the meaning under discussion.

I also do not consider constructions with nominal classifiers in this study. Nominal classifiers are adnominal elements the main function of which is to indicate a property of a head noun, according to which it can be attributed to a more extensive class of objects (animals, women, weapons, etc.) (Grienvald 2000: 64-68). These elements can be optional, and in some cases, a name may choose different classifiers depending on the context. Some systems of nominal classifiers are very limited and closed, while others may have hundreds of elements and no clear boundaries. That practically means that in some languages there is no clear difference between lexical “sortal” nouns and such purely functional elements as classifiers (Aikhenvald 2000: 81-92). Whenever there is a doubt about the free status of a generic term, the example is excluded.

Since the information on intonation integrity, which is the main defining feature of close apposition, is generally not presesnt in written grammars, I rely on the presence or absence of commas in the translation given by the author of the grammar in differentiating close appositions from loose appostions.



3. Expression of specification in the languages of the sample

This chapter introduces the inventory of \ morphosyntactic means used to code relationships between the members of constructions of specification in the languages of the world. Below I consider cases of simple juxtaposition of nouns (apposition in its strict syntactic sense) (3.1.1), coding by means of subordination (3.1.2), and examples of less common special strategies observed sporadically in the languages of the sample. The material of this chapter demonstrates that from a typological point of view, relation of specification is coded in a pretty universal manner, although the observed variation is revealing in terms of the nature of the constructions of this sort.

3.1 Juxtapositional strategy and its syntactic properties

Juxtaposition is the most common strategy for expression of specification in the surveyed sample. It is observed in all of the considered languages (perhaps, except for 5 -- Movima, Niuean, Luganda, Ma'di and Adyghe, all discussed below) and is the only way of expressing specification by a proper name in 65 out of 95 languages of the sample. It is found in all the macro-areas and almost all the language families. Examples of constructions of this kind are provided below for Evenki (7), Chukchi (8), Tundra Nenets (9) (Eurasia), Hausa (10), Irakw (11), Sango (12), Mende (13), Supyire (14) (Africa), Popoluca (15), Seri (16), Northwest Sahaptin (21) (North America), Ese Ejja (17), Chácobo (17), Cocama-Cocamilla (18), Huambisa (19), Tariana (20) (South America), Taba (22) and Gayo (23) (New Guinea) and many others:

(7) Evenki (< Tungusic< Altaic, Lenin i Stalin, glosses are mine)

à. ãåíåðàë-ve Íîñîâè÷-va

general-acc Nosoviè-acc

`general Nosoviè'

b. gorod-tu Omskaj-du

gorod-dat Omsk-dat

`to the city of Omsk'

(8) Chikchi (< Northern Chukotko-Kamchatkan < Chukotko-Kamchatkan)

a. Ìàé?ûíûì-ãûïû Íàçàðåò (Kniga proroka Iony, glosses are mine)

city-abl Nazareth

`from the city of Nazareth'

b. Âààìãû-ïû Àíàäûðãû-ïû (N. Zorina (personal communication), glosses are mine)

river-abl Anadyr-abl

`[we left] the river of Anadyr'

(9) Tundra Nenets (< Samoyedic < Uralic, (Nikolaeva 2014: 171))

a. s?wa n'a-m'i Wera-m'i

good friend-1sg Wera-1sg

`my good friend Wera'

b. Moskva m?r

Moscow city

`the city of Moscow'

(10) Hausa (< West Chadic < Afro-Asiatic, (Jagger 2001: 41), glosses are mine)

a. ya?rinya?Ì-ta? Jummai

girlfriend/1sg-1sg Jummai

`my girlfriend Jummai'

b. aÌbo?ka-nmuÌ Mu?sa? daÌ MammaÌn

friend-1pl Musa and Mamman

`our friends Musa and Mamman'

(11) Irakw (< Southern Cushitic < Afro-Asiatic, (Mous 2007: 27)

mulqumo-?eìe? martiiìni

friend-1.sg.poss Martin

`my friend Martin'

(12) Sango (< Ubangi < Niger-Congo, (Samarin 1963: 44), glosses are mine)

fadeì m? wara {m?ìreng?ì tiì mbi Wanzaka Louise} The curly brackets are provided by the author.

fut you find clild poss 1sg W. L.

`You will find my child Wazaka Louise'.

(13) Mende (< Western Mande < Mande, (Hoel et al. 1994: 81)

korina sanje si-hi tekla, hulayi si-hi moses, jikisi fri-hi

Korina older_sibling 3s.f-poss tekla man 3s.f-poss Moses child 2d.-poss

peter nella roberta

Peter Nella Roberta

`Korina's older sister Tekla and her husband Mose's children are: Peter, Nella and Roberta'.

(14) Supyire ( < Gur < Niger-Congo, (Carlson 1990: 349))

na cevoo zhyeì, taaì mu na ?-k?ì?ìg?ì ke

my.nondecl friend Zieì where you prog ip-go.impfv locq

`My friend Zieì, where are you going?'.

(15) Popoluca (Sierra) (Mixe-Zoque < Mixe-Zoque, (Boudreault 2009: 263, 386))

a. ?i+winykej?a?y ta?+naaba+tam Malia

?i+winy=kej-?a?y-W tan+?aapa+tam Malia

3erg+*face=appear-ben-cmp ipsr+mother+plu_hum Maria

`Our mother Mary appeared'.

b. jem?k ??ch ?an+t?km?? ?aattep?et

jem?k ??ch ?an+t?k=m? ?aattep?et

there 1pro xpsr+house=in Soteapan

`There in my town Soteapan'.

(16) Seri (< Seri < Hokan, ( Marlett, Moser 2004: 475, 547)

a. René quih (cmaacoj) ai José Ángel

the.fl old.man 3p-father.me

`René's father José Ángel'

b. heezitim Betánia hapáh quij heezitim Jerosalém hapáh quij

city sn-pv-say the.cm city sn-pv-say the.cm

hantaxal xah cöquiij iha.

near+mod atten 3io-sn-sit d_cl

`Bethany is near Jerusalem'.

(17) Ese Ejja (< Tacanan < Panoan, (Vuillermet 2012: 356, 675))

a. Ma yawa-nei-nei=pa yowa=a ma presidente

then long_ago-very-red=rep whatchamacallit=erg dem presidente(Sp)

German Bush=ya meshi kya-ka-a=pwa campesino.

German Bush=erg land give-3a-rpas=rpas indigenous(Sp)

`This, long ago, when -- what's his name again -- the President German Bush gave lands

to the indigenous people so that they (can) work their fields'.

b. Oya ye<kwaya>ka-a=kwana=pwa ba'eña-a

3abs go_downriver<KWAYA>-pras=pl=pras land-mot_purp

Madidi kwei=xe Barracon=jo ba'eña-a=pwa

Madidi river=perl b.=loc bearth-pras=pras

`They left to land on the river Madidi, they landed at Barracon.'

(17) Chácobo (Panoan < Panoan, (Tallman 2018: 1339), Bible in Chácobo)

a. ?waìti Yoni=yaì

grand_mo Yoni=com

`with the grandmother Yoni'

b. yaca Nazaret

city Nazareth

`the city of Nazareth'

(18) Cocama-Cocamilla (< Tupi-Guaraní < Tupian (Yopán 2010: 770)

?wati=tsui ts=umi lima ritama=pura

high=abl 1sg.f=see Lima village=foc

`From above I see Lima city'.

(19) Huambisa (Jivaroan < Jivaroan (Pen?a 2015: 333, 683, 780))

iat?su-ruì haimito mina iaakata Soledad

brother-1sg Jaime (dim) 1sg.gen town Soledad

ii=ka himaraì w?-sa-ta-hi

1pl=foc two\iter go-att-imm.fut-1pl.sbj+decl

`My brother Jaime, we (i.e. you and I) will go two times to my town Soledad.'

(20) Tariana (< Northern Maipuran < Arawakan, (Aikhenvald 2003: 480, 630))

a. nu-we-ri Jovino

1sg-younger.brother-masc Jovino

`my younger brother Jovino'

b. rio Orinoco-naku

river Orinoco-top/non.a/s

`river Orinoco'

(21) Northwest Sahaptin (< Sahaptian < Penutian, (Jansen 2010: 290))

wiìyat-pa niìxyaìawi-pa p?ìt'x?anuk-pa

far.away-loc Nixyaawi-loc mountain-loc

`far away at the Nixyaawi Mountains'

(22) Taba (< South Halmahera-West New Guinea <Austronesian (Bowden 1997: 208)):

kampung Kota

village Kota

`Kota village'

(23) Gayo (Northwest Sumatra-Barrier Islands < Austronesian, (Eades 2005: 137, 201, 223))

a. engi=eì, peteri Meleìlacanu

younger.sibling=3.poss princess Meleìlacanu

`her ypunger sister, princess Meleìlacanu'

b. serdadu ni Belene si ara i kute TakeÌngen

`soldier poss Dutch rel exist loc town Takengon

`the Dutch soldiers who are in the city of Takengon'

c. urang-urang kampung Lumut [be-doa] [munatang-an pumu ku atas]

red-person village Lumut mid-pray ao:rise-caus1 hand to top

`The people of Lumut village prayed, raising their hands up'.

The problem in giving any analysis to the structures above and similar examples from the other languages is that description of apposition, if at all present, is very scarce. Authors usually give no account of the internal and external syntax of appositive constructions, and there are few special articles providing an analysis of apposition in a particular language.