Serial Verbs

This article analyses the grammaticalization of the pre-verbal particle kom (from the English verb " to come ") in Nigerian Pidgin (NigP). Given that the construction kom + V has probably born out of a context of verb serialization, we will start with a definition of serial verb constructions (SVC) and also provide a panorama of the SVCs in NigP. Serial verb constructions have triggered many cases of grammaticalization in NigP. Later, we will discuss the concept of " grammaticalization " and apply it to the field of contact linguistics: in creoles one can often observe complex constructions that seem instances of grammaticalization but they do not really represent an internal development of the creole. Some grammaticalized items actually mirror diachronic development which took place in the substrate languages, the speakers' first languages. The analysis of pre-verbal kom in NigP will be proposed as an example of ordinary grammaticalization, a diachronic process of functionalization in fieri, triggered by similar structures in the substrates but finally language-internal.

Questo articolo sarà dedicato all'analisi di un caso di grammaticalizzazione in Nigerian Pidgin (NigP nel testo), quella di kom (dall'inglese " to come " : venire) usato come particella pre-verbale non lessicale. Siccome la costruzione kom + V è probabilmente nata come un caso di serializzazione verbale, si partirà da una definizione delle costruzioni a verbi seriali (SVC nel testo) e da una panoramica di questo tipo di strutture in NigP. Le costruzioni a verbi seriali hanno dato luogo a diversi fenomeni di grammaticalizzazione in NigP. Si procederà, dunque, con la discussione di quanto le teorie tradizionali della grammaticalizzazione possano prestarsi all'analisi delle lingue di contatto: nei creoli, infatti, si osservano spesso fenomeni di calco di strutture già grammaticalizzate presenti nelle prime lingue dei parlanti, le lingue di sostrato, così da escludere, in tali casi, processi di sviluppo interno ai creoli. L'analisi del kom pre-verbale in NigP, sarà portata come esempio putativo di grammaticalizzazione ordinaria, un processo di funzionalizzazione diacronico (in fieri) del tutto interno alla lingua. La particella pre-verbale kom in Nigerian Pidgin

Barayin is an East Chadic language spoken by around 5000 people in the Guera region of the Republic of Chad. This dissertation examines a particular type of syntactic construction in the language, serial verb constructions, from the perspectives of typological (or comparative) syntax, descriptive grammar, and the formal syntactic theory of Lexical-Functional Grammar (LFG). Typologically, serial verb constructions are problematic because they represent a heterogeneous set of multiverb constructions that have features that do not fit into traditional syntactic categories like subordination, conjunction and adjunction. Part A of this dissertation describes these problematic features in detail, providing a succinct overview of the literature which can serve as a resource for field linguists describing similar constructions. Part B of the dissertation gives a detailed description of the morphology, syntax and semantics of serial verb constructions in Barayin. These chapters contribute to our knowledge of the world’s languages by documenting a complex syntactic phenomenon in an area of the world where most of the languages are significantly understudied. The most common type of SVC in Barayin involves a deictic motion verb. The motion is normally (but not always) understood to take place prior to the activity or state predicated by the main verb. The formal analysis of Barayin SVCs in Lexical-Functional Grammar in Part C uses recent developments in the theory to show how argument sharing in SVCs can be represented in a connected s-structure that conforms to the standard mechanisms of LFG. The approach is compared to two previous analyses of complex motion predicates in other languages which appeal to a non-standard formal mechanism to model complex predicates.

This thesis provides an initial descriptive analysis of a largely undescribed Austroasiatic language of Meghalaya, northeast India. Pnar has often been overlooked because of its lexical similarity to Khasi, with which it shares a common identity and society. Features such as verb-initial constituent order, processes of nominalization and derivation, prepositional marking, a rich gender system, a large set of deictic markers, and its use of plural marking within classifier phrases prove that Pnar is richly deserving of study. The following pages attempt to thoroughly examine the phonological, morphological, and grammatical structure of this language for the benefit of the speakers and the larger linguistic community. Appendices include a selection of texts, a list of elaborate expressions, and a 1,600 word lexicon.

Dedicated url: http://hdl.handle.net/10356/62519

This paper continues a whole series of previous articles on the so-called Russian double verbs, but examines them for the first time on the basis of data from the Russian national corpus. It focuses on just one grammatical form out of the whole paradigm, viz. imperatives in the 2pl. Among the topics
scrutinized and quantified are: the ratio of juxtaposition vs distant position (with one dividing element) of both verbs and of prototypical (asymmetric) vs non prototypical (symmetric) pairs, the inversion of V1 and V2, the scope of negation, pairs with mixed verbal aspect, the desemanticisation of V1, the
historical rise of some V1 and decline of others. The results obtained by this research corroborate the overall picture drawn in our previous works, but provide significant new insights such as the overwhelming predominance of the prototype over symmetric pairs, exclusive negation of V2, prevailing distribution of ipf V1 + pf V2 in aspectually mixed pairs, etc.

Uncorrected proofs, Oct 2020. To appear in DeGruyter Mouton series Empirical Approaches to Linguistic Typology collection from the ALT 2017 workshop on Associated Motion

This chapter investigates the expression of associated motion and directional motion in the form of serial verb constructions (SVCs). In a sample of 124 languages with SVCs, 80% have motion SVCs. The most common types are directional SVCs, in which a path-of-motion verb combines with another motion verb, and prior associated motion SVCs expressing motion prior to the activity or state predicated by the other verb in the construction. Concurrent motion and subsequent motion are much less common. In a prior motion SVC, the motion verb nearly always precedes the other verb, and the figure on the path of motion is the subject. In a directional SVC, the path-of-motion verb nearly always follows the other verb, and the grammatical function of the figure on the path of motion can vary according to the semantics of the main verb in the construction.

NB: For a revised and condensed version of this overview, see the article Serial Verb Constructions in Annual Review of Linguistics 7 (2021).
http://www.annualreviews.org/eprint/ZK6GFQCYSQDNXK2WXI7B/full/10.1146/annurev-linguistics-031920-115317

This paper was a draft of an extended overview of the literature on serial verb constructions prepared as part of the background reading for my PhD dissertation. Feel free to use this document to search for other references, but be aware that there may be some errors or misrepresentations that have not been corrected. Please do not cite or quote this document.

Since the 1970s, serial verb constructions (SVCs) have been discussed widely in African, Oceanic and many other languages in different parts of the world. This paper gives an overview of the most important generalizations about serial verb constructions that have been proposed and that do seem to hold if a sufficiently restrictive definition of the concept is adopted. The main problem of the earlier comparative literature is that the notion of a SVC has not been delimited clearly, and/or has been formulated in much too wide terms. This can be seen as a result of the confusion between comparative concepts and natural kinds: Serial verb constructions have (most often implicitly) been regarded as natural kinds (universal categories), so that phenomena in additional languages were regarded as SVCs even when they had somewhat different properties. This procedure inevitably leads to a fuzzy and very broad understanding of the concept, with a prototype structure that does not allow falsifiable claims. Here I propose a narrow definition of SVC and formulate ten universals that are apparently true of all serial verb constructions in this narrow sense.

In this dissertation a methodology for identifying and analyzing serial verb constructions (SVCs) is developed, and its application is exemplified through an analysis of SVCs in Koro, an Oceanic language of Papua New Guinea. SVCs involve two main verbs that form a single predicate and share at least one of their arguments. In addition, they have shared values for tense, aspect, and mood, and they denote a single event. The unique syntactic and semantic properties of SVCs present a number of theoretical challenges, and thus they have invited great interest from syntacticians and typologists alike. But characterizing the nature of SVCs and making generalizations about the typology of serializing languages has proven difficult. There is still debate about both the surface properties of SVCs and their underlying syntactic structure. The current work addresses some of these issues by approaching serialization from two angles: the typological and the language-specific.

On the typological front, it refines the definition of `SVC' and develops a principled set of cross-linguistically applicable diagnostics. From the existing set of surface properties, four core characteristics are distilled: main verbhood, monoclausality, single eventhood, and argument sharing. A construction must have all of these properties in order to qualify as an SVC. Once these underlying semantic and syntactic properties of SVCs are identified, a detailed and explicit set of criteria is developed that allows these underlying properties to be tested in any language.

The latter part of the dissertation offers a case study in the use of these diagnostic criteria by applying them to multi-verb constructions in Koro. Testing these constructions against the definition of SVCs developed in the dissertation reveals that although there are numerous multi-verb constructions in Koro that appear to fulfill the surface criteria for SVCs, only one of these can be considered a true SVC. This construction has a VP-shell structure, in which V1 is a path or locative verb that takes V2 as its complement. The shared argument is the subject of V2, providing a counter-example to Baker's (1989) claim that SVCs obligatorily share an internal argument. Constructions that instead involve adjunction of V2 to V1 are shown through detailed semantic investigation to be disqualified as SVCs, because they do not exhibit the expected entailments. This is surprising because they superficially resemble proto-typical SVCs. The syntactic and semantic analysis of these constructions leads to the hypothesis that true SVCs must have a relation of complementation between the verbs, while adjoined or coordinated constructions cannot be considered SVCs.

This is a grouped bibliography of literature on complex verb constructions (serial verbs, coverbs, complex predicates, etc.). Please feel free to use it, comment on it, or send me feedback of anything I might have missed. If you make extensive use of my list for academic publication, I would appreciate an acknowledgement.

Diller kendisini oluşturan iç unsurlarının birbirleriyle olan ilişkilerini bir mekanizma gibi çalıştırarak varlıkları, kavramları ya da bu varlıkların hareketlerini ifade ederler. Bir başka deyişle ses, biçim, kelime ya da söz dizimi çeşitli birlikteliklerle eşgüdümlü biçimde hareket ederek anlamın ortaya çıkmasını sağlarlar. Dilbilgisi dallarının birlikte işlev görmesiyle ortaya çıkmış yapılardan biri de seri fiil yapılarıdır. Seri fiil yapıları, birden fazla fiilin yan yana gelerek ya da tek bir fiilin birden fazla hareketi içererek bütüncül bir bakışla "tek bir olayı" işaretlediği fiilli yapıları ifade etmektedir. Seri fiil yapılarında tek bir fiil içerisinde iki ayrı fiille ifade edilebilen ve doğal olarak gerçekleşen ancak "tek olayı" işaretleyen yapıların yanı sıra en az iki fiilin söz dizimsel olarak bir araya gelmesiyle "tek bir olayı" işaretlemesi önem taşımaktadır. XIX. yüzyılda yapılan ilk çalışmalarda yalnızca Afrika dillerinde bulunduğu düşünülen seri fiil yapılarının zamanla diğer bilimsel çalışmalarda dünyadaki birçok dilde de yer aldığı görüşü hâkim olmuştur. Türkçede seri fiil yapılarıyla ilgili birtakım çalışmalar mevcuttur. Bu çalışmada seri fiil yapılarıyla ilgili yapılmış çalışmalarda ortaya koyulan görüşlerden hareketle Türkçede bu yapıların görünümü artzamanlı ve eşzamanlı olarak incelenmiştir. Bununla birlikte Türkçe seri fiil yapılarının özelliklerinin belirlenmesi amaçlanmıştır.

ABSTRACT Languages act as a mechanism of the relations of the internal elements that build them up, and express their existence, concepts, or movements. In other words, phonetics, morphology, lexicology, and syntax move in a coordinated manner with the various associations in order to provide meaning. One of the structures that emerged as a result of the collaboration of grammatical categories is the serial verb structure (SVC). Serial verb structures refer to actual structures in which more than one verb come together or a single verb, thus implying a single event with a single view, including multiple movements marking a single event with a holistic look. Moreover, is important to mark a single event with a syntactic combination of at least two verbs. Studies made in the 19th century frequently claimed the serial verb structures existed in specific African languages, however studies on many other languages/language families were done as well. With this in mind, a number of studies in Turkish about serial verb structures have been put forth. This study both synchronically and diachronically examines the structure of SVCs in Turkish based the opinions from aforementioned studies, as well as aims to determine their characteristics.

The two English serial verb constructions, go-VERB and come-VERB, are subject to a peculiar morphological restriction, called the Bare Stem Condition. The constraint states that neither the verb in V1 nor its directly adjacent dependent verb in V2 can (or does) occur with (overt) inflection (<i>Go see a doctor!</i>, They invited him to come visit them, but *She goes sees a doctor, *They were going playing golf). The constraint is near-unique to English; it is non-predictable from other rules or patterns of English grammar and without any obvious systemic motivation. Due to its apparent arbitrariness, the constraint has mainly been approached from formal and parametric perspectives.
By contrast, the current approach takes a usage-based view and argues that the constraint is functionally conditioned by the constructions' semantics and the make-up of the English morphological paradigm. On the one hand, the constructions show a notable preference for directive-commissive functions (orders, commands, invitations, recommendations, suggestions, etc.), arguing that go/come-VERB have a hortative schema core. On the other hand, such non-assertive propositions in English are encoded using bare verb forms (imperatives, infinitival complements). The skew towards bare form, albeit motivated, is thus seen as an (historical) accident, which is contemporarily highly conventionalized and entrenched.
Methodologically, the approach suggests a syntactic co-location approach (“You shall know a construction by the company it keeps”), inferring constructional semantics by distributional preferences of co-occurrence with leftward syntactic units (e.g., imperatives, (semi)modals, let's-constructions, speech act verbs). The usage of go/come-VERB is investigated in four case studies of constraint violation, network relations (viz. go/come-and-VERB), diachronic development, and schema acquisition. The evidence suggests that an explanatory attempt of a seemingly arbitrary formal constraint is possible without recourse to innate linguistic knowledge.

This thesis focuses on nominalization of serial verb constructions (SVCs) in the Akan language. The study develops a relevant typology of serial verb nominalization on the basis of semantic integration and lexicalization using a prototype theory (PT) framework. The three degrees of semantic integration for serial verbs in Akan are Full Lexicalized-Integrated Serial Verb Constructions (FL-ISVCs), Partial Lexicalized-Integrated Serial Verb Constructions (PL-ISVCs) and Clause Chaining Serial Constructions (CCs or CSCs). Each type of SVC is analyzed on the basis of how it is nominalized, the degree to which nominalization occurs and whether nominalization can occur at all. Various corpuses were consulted in three major literary dialects of Akan: Asante Twi, Fante and Akuapem Twi. Further, native speakers of each of these dialects were consulted to ascertain the goodness of various attested serial verb nominals (SVNs) in Akan.

Because Full Lexicalized-Integrated Serial Verb Constructions behave as lexicalized idioms, four criteria of idiomaticity are applied to them including collocability, familiarity, flexibility and compositionality (Barkema 1996). The results from the study show that over 98% of all FL-ISVCs identified have nominal counterparts while less than 3% of all PL-ISVCs identified have nominal counterparts. CSCs seem to nominalize haphazardly as frozen sentences, proverbs, idioms and figures of speech which primarily function as denotata and designata. While there was some degree of interdialectal variability with regard to individual SVNs, the pattern of nominalization behavior on the basis of degrees of semantic integration and lexicalization remained consistent across dialects.

Many Australian languages form complex predicate (CP) constructions where one (or more) uninflecting open-classed coverb combines with an inflecting verb (IV) belonging to a closed class. It has been argued that these types of CPs constitute an areal feature structurally distinct from other CP constructions such as light verb constructions (McGregor, 2002; Schultze-Berndt, 2000).
Recently, some authors (Bowern, 2014; Hoffmann, 2014; Nordlinger, 2014; Reid, 2002) have focused their at- tention on an additional type of multi-verb construction, namely serial verb constructions in Ngan’gityemerri, Wambaya, and Jingulu and serial coverb constructions in MalakMalak, Wagiman, Kamu, and Matngele. This paper aims to provide a systematic overview of the types of serial verb constructions that occur in the Daly River region of Australia (and beyond). I explore how the nexus between the coverb and the IV, a language’s ability to form semi-independent predicates without IVs, and the number of IVs in a language are related to the type of serialization a language allows.

Practitioners of syntactic reconstruction have not acknowledged that arbitrariness and iconicity influence syntactic change, and that they therefore need to be incorporated into methods of reconstruction. I argue that iconicity creates a directional tendency in syntactic change, privileging structures that are more iconic. I propose a method for incorporating this fact into methods of syntactic reconstruction. I demonstrate the application of this method on two pieces of reconstructed syntax: orientation serial verb constructions and left-peripheral topics. Both case studies are from Proto-Sogeram, the ancestor to ten languages of Papua New Guinea. A third, briefer case study concerns Proto-Carib.

Sgaw Karen is a Sino-Tibetan language spoken in Burma by about 1,284,000 people. It is largely analytic/isolating, uses contrastive tones, and has a canonical SVO word order. An interesting issue in the language is the mechanics of its complex verb phrases, particularly serial verbs.
Serial verb constructions, when a single clause contains multiple verbs not joined by a conjunction or other subordination, are common in Sgaw Karen. Some serial verbs contain two (or more) verbs with similar valency, as in (1) where both 'fall' and 'run' are intransitive, or (2) where both 'take' and 'get' are transitive. (CL = Classifier.)

1) jə¹ lɔ⁶ xʷe²
1.sg fall run
"I fall."

(2) ʔa¹wɛ¹ hi⁴ ne² tʰwi⁵ tə¹ dɨ¹
3.sg take get dog one CL
"He got a dog."

However, many serial verbs n Sgaw Karen contain mismatched valencies, as in (3), where 'cut' is transitive but 'fall' is intransitive.

3) jə¹ ku⁵ lɔ⁶ θe⁵
1.sg cut fall tree
"I cut down a tree," or "I cut a tree; the tree fell."

These data are problematic for a number of reasons. Both verbs have theta-roles to assign, for a total of three (i.e. Agent and Patient of 'cut' and Patient of 'fall'), but only two arguments exist in the sentence. Under classic syntactic theories, every DP argument in a sentence can only receive one theta-role. Additionally, VP structures do not allow for multiple verbs in many theories, and both the underlying structure of the serial verb and the assignment of a single tense to multiple verbs are problematic for these theories.
I establish the limits on combining differing valencies, determining whether verbs can freely combine or whether constraints on serial verbs exist. I determine the possible directions of change, whether an increase or decrease in valency is more common in Sgaw Karen. Additionally, I address similar complex verb phrases that involve changes in valency.
Further, I explore various solutions to this issue, evaluating different posited structures for this class of valence-changing constructions including verb compounding, verbs as modifiers of a head verb, and a Minimalist Syntax approach.

My 2018 BA thesis devoted to the preliminary typology of Verb-Verb compounds. I study the order of semantic Heads and Modifiers in three types of verb-verb compounds which are also defined semantically. I argue that while in Mental process compounds the order of components correlates with the general branching direction of a given language, the component order in compounds involving motion verbs is determined by the distinction between Path verbs and other verbs. This thesis also contains some case studies devoted to the grammaticalization of verb-verb compounds.

The objective of the research is to classify the serial-verb constructions in Thai automatically by using the word classes from Thai WordNet to classify verbs in the sentence. Due to the Thai language has the extend-to-the-right structure and put the adjective after the noun. Its overall grammar characteristic is the &quot;Subject-Verb-Object&quot; or SVO type. And Thai language can be communicated using one verb after another within the same sentence, that we called &quot;Serial Verb&quot;. Today we already have many researches about this serial-verb constructions, but no research is about its automatic classification

The author takes up the “V1 conjunction wa V2” complex verb construction in Ainu, an endangered language spoken in north Japan that is genetically and typologically unrelated to Japanese. Assuming that the “V1 conjunction wa V2” construction corresponds to the “V -te V” complex predicate construction in Japanese, the author exploits diagnostic tests comparable to those applied for Japanese V -te V complexes to see whether the Ainu construction makes up one predicate or two, leading to nine  subtypes of the construction. By comparing Hokkaido Ainu with their Japanese V1 -te V2 counterparts, she suggests that six subtypes with V2s meaning EXIST, COME, GO, SEE, GIVE, PUT, but not DISAPPEAR, FINISH, or HEAR, are attested in both languages and speculates that some of the shared subtypes may have developed in Hokkaido Ainu under the influence of Japanese.

The objective of the research is to classify the serial-verb constructions in Thai automatically by using the word classes from Thai WordNet to classify verbs in the sentence. Due to the Thai language has the extend-to-the-right  structure  and  put  the  adjective  after the  noun.  Its  overall  grammar  characteristic  is  the "Subject-Verb-Object" or SVO type. And Thai language can be communicated using one verb after another within the same sentence, that we called "Serial Verb". Today we already have many researches about this serial-verb constructions, but no research is about its automatic classification

In this paper, serialized chains of verb forms of the type pojdem posmotrim are first examined with respect to their grammatical homonymy (1PL.IMP vs. FUT1PL) and lexical, syntactic and contextual cues that allow to disambiguate it. In section 2, a thorough analysis of data from the Russian national corpus serves to capture the aspectual and lexical constraints of such chains. The last section contains a systematic comparison of said forms with serialized imperatives in the 2nd pl.

In the typology of motion lexicalization, two types of languages have traditionally been distinguished: satellite-framed and verb-framed. Serializing languages are difficult to fit into this typology and have been claimed to belong to a third type: equipollently framed. In this paper I use grammatical criteria to show that Avatime, a serializing language, should indeed be classified as equipollently framed. I also study motion descriptions in narratives. Avatime is similar to other serializing languages with respect to path elaboration, but unlike other serializing languages it has low manner saliency. Equipollently framed languages thus do not behave as a single type.

In this paper I will demonstrate how a series of micro changes are related within the synchronic gradience of the item take in Western metropolitan varieties of Nigerian Pidgin (NigP). I will show that take functions as a main verb, as an instrumental serial verb and as an incipient modal verb. While the presence of serial verb constructions (SVCs) can be attributed to a calque from the Kwa/Benue-Congo substrates, I will make an argument for an internal and gradual path of grammaticalisation from serial to modal take, due to the re-bracketing of the collocation take+V2 as a non-serial VP. I will support my claims with an analysis of take’s occurrences in a corpus of NigP collected in Lagos in 2007.

Akan verb-verb nominal compounds exhibit unusual formal and semantic properties, including extreme formal exocentricity, where the composition of two verbs yields a noun some of whose semantic properties may not be directly coded in the constituents, and argument structure suppression, where no argument of either constituent can occur in the compound. The purpose of this paper is two-fold. First, I delineate the membership of the class, showing that some of the constructions listed in the literature as verb-verb compounds do not belong to the class; they have formal features that betray them as affix-derived nominals. Secondly, I discuss the rather idiosyncratic properties of the compound. I argue that the form class is inherited from a metaschema for compounding in Akan which bears a nominal output category. Again, it is a unique constructional property of Akan verb-verb compounds that, unlike other verb-involved compounds, they do not allow any argument of the constituents to become part of the compound. These extra-compositional holistic properties can be accounted for straightforwardly in a framework like Construction Morphology which does not assume that every property in a construction must emanate from its constituents. This study provides evidence for the view that constructions can have holistic properties.

This paper gives an ac­count of Se­ri­al Verb Con­struc­tions (SVCs) in Man­darin Chi­nese. After a ty­po­log­i­cal pre­sen­ta­tion of the phe­nomenon, we give an overview of the Chi­nese data. The in­ven­to­ry of SVC types is clas­si­fied ac­cord­ing to causal and tem­po­ral re­la­tions be­tween the com­po­nents. We dis­cuss prag­mat­ic con­di­tions on the use of SVCs and al­ter­na­tive, se­man­ti­cal­ly equiv­a­lent con­struc­tions. An HPSG anal­y­sis is pro­posed for marked SVCs which uses the in­ter­ac­tion be­tween as­pect mark­ing and the set of pos­si­ble sub­or­di­na­tive re­la­tions to de­duce the ex­tra-​lex­i­cal mean­ing of the con­struc­tion. Par­tic­u­lar at­ten­tion is paid to the syn­tac­ti­cal­ly pe­cu­liar SVC with shared in­ter­nal ar­gu­ments, which is ac­count­ed for by a non-​can­cel­la­tion ap­proach to va­lence re­quire­ments.

Multi-verb constructions are a cross-linguistically widespread phenomenon. They share features, but also display systematic differences, the natures of which have been discussed widely (Amberber et al., 2007; Aikhenvald, 2006; Bowern, 2014; Crowley, 2002; Durie, 1997; Foley and Olsen, 1985,?; Schultze-Berndt, 2000). MalakMalak and Matngele, two head-marking, polysynthetic non-Pama-Nyungan Daly languages of Northern Australia provide an opportunity to investigate several multi-verb constructions, specifically, complex predi- cates (CPs), serial coverb constructions (SCCs), and compound coverbs (CCs), within a single language. In this paper I systematically test for negation, prosodic word status, subordination, semantic compositionality, and syntactic constituency to reveal distinguishing and combining features of these multiverb constructions.

Data from instrumental serial verb constructions (SVCs) in Malayo-Polynesian languages (e.g. Ambonese Malay: Paauw 2008; Colloquial Jakartan Indonesian: Sneddon 2006; Kupang Malay: Jacob and Grimes 2011; Taba: Bowden 2001) support the hypothesis that a pragmatic principle of temporal iconicity prevents the grammaticalization of anti-iconic fixed verb orders in instrumental SVCs. The same data falsify the competing hypothesis, that verb order is constrained indirectly by a “thematic hierarchy”.

It has been observed that in instrumental SVCs which use the verb ‘take’ to introduce the instrument the ‘take’ verb always precedes the main verb (Muysken 1988; Sebba 1987:144-145). This is true regardless of whether the language is right-headed (VO) or left-headed (OV). See the examples of an instrumental SVC in Sranan (VO), example 1, and an instrumental SVC in Ijo (OV), example 2. The relationship between the verbs in these SVCs does not correspond to the verb-complement relationship.

Sranan (Surinamese creole) (Muysken 1988, cited in Carstens 2002; cf. Muysken and Veenstra 1994)

(1) no teki baskita tyari watra
no take basket carry water
Don’t carry water with a basket!

Ijo (Niger-Congo) (Williamson 1965:53)

(2) áràú̩, zu ye áki̩, buru tèri-mí
3SG basket take yam cover-PST
She covered the yam with a basket.

Several linguists have proposed that the order of verbs in SVCs, including instrumental SVCs, can be explained by appealing to a principle of temporal iconicity (Durie 1997; Li 1993; Lord 1993; Muysken 1988; Nishiyama 1998; Tai 1985). The order of the verbs reflects the chronological order of the parts of the predication that each verb is associated with. In instrumental ‘take’ SVCs, the verb ‘take’ occurs before the main verb in linear order because the transfer of possession it represents precedes the activity of the main verb chronologically. An objection to this view is that when the activities associated with each verb are simultaneous, verb order often remains fixed, despite the lack of temporal sequence (Baker 1989:536). For example, instrumental SVCs that employ the verb ‘use’ (instead of ‘take’) to introduce the instrument argument, show two possible word orders. In some languages, the ‘use’ verb occurs first, as in example 3. In other languages, ‘use’ occurs second, as in example 4. Neither word order violates temporal iconicity, since the activities are simultaneous (Bowden 2001:313). However, temporal iconicity does not explain why alternating word order in a single language has not been reported. The answer to this objection is that temporal iconicity is not a synchronic constraint on SVCs, but part of a diachronic explanation for why anti-iconic verb orders do not become grammaticalized (Good 2003:437, 444).

Thai (Filbeck 1975:120)

(3) sùk cháy phráa khôon tônmáy
Sook use machete cut tree
Sook chopped down the tree with a machete.

Taba (Malayo-Polynesia) (Bowden 2001:299-300)

(4) n=pun bobay n=pake sandal
3SG=kill mosquito 3SG=use thong
He killed the mosquito with a thong.

The same data disprove the competing explanation of verb order in instrumental SVCs. Baker (1989) and Carstens (2002) argue that the verb order patterns should be explained by appealing to thematic roles. Their approach predicts that, in an instrumental SVC, the verb assigning the instrumental role will always follow the instrument argument. Data of ‘use’ instrumental SVCs from several Malayo-Polynesian languages (including example 4) falsify the prediction of the thematic hierarchy analysis of instrumental SVCs. The ‘use’ verb and its argument (the instrument) follow the main verb. Verb order in instrumental SVCs cannot be explained by appealing to a constraint on the linear order of arguments with certain thematic roles.

References

Baker, Mark (1989). Object sharing and projection in serial verb constructions. Linguistic Inquiry, 20(4):513–553.
Bowden, John (2001). Taba: description of a South Halmahera language. Pacific Linguistics.
Carstens, Vicki (2002). Antisymmetry and word order in serial constructions. Language, 78(1):3–50.
Durie, Mark (1997). Grammatical structures in verb serialization. In Alsina, Alex, Bresnan, Joan, and Sells, Peter, editors, Complex predicates, pages 289–354. CSLI Publications, Stanford.
Filbeck, David (1975). A grammar of verb serialization in Thai. In Harris, JG and Chamberlain, JR, editors, Studies in Tai linguistics in honor of William J. Gedney, pages 112–129.
Good, Jeffrey (2003). Strong linearity: Three case studies towards a theory of morphosyntactic templatic constructions. PhD thesis, University of California, Berkeley.
Jacob, June A and Grimes, Charles E (2011). Aspect and directionality in Kupang Malay serial verb constructions: Calquing on the grammars of substrate languages. Creoles, their substrates, and language typology, pages 337–366.
Li, Yafei (1993). Structural head and aspectuality. Language, 69(3):480–504.
Lord, Carol (1993). Historical change in serial verb constructions, volume 26 of Typological studies in language. John Benjamins, Amsterdam/Philadelphia.
Muysken, Pieter (1988). Parameters for serial verbs. In Manfredi, Victor, editor, Niger-Congo syntax and semantics 1, pages 65–75. Boston University African Studies Center, Boston.
Muysken, Pieter and Veenstra, Tonjes (1994). Serial verbs. In Arends, Jacques, Muysken, Pieter, and Smith, Norcal, editors, Pidgins and Creoles: An introduction, pages 289–302. John Benjamins.
Nishiyama, Kunio (1998). V-V compounds as serialization. Journal of East Asian Linguistics, 7(3):175–217.
Paauw, Scott H (2008). The Malay contact varieties of eastern Indonesia: A typological comparison. PhD thesis, State University of New York at Buffalo.
Sebba, Mark (1987). The syntax of serial verbs: An investigation into serialisation in Sranan and other languages. J. Benjamins, Amsterdam; Philadelphia.
Sneddon, James (2006). Colloquial Jakaratan Indonesian. Pacific Linguistics.
Tai, James H.-Y. (1985). Temporal sequence and Chinese word order. In Haiman, John, editor, Iconicity in Syntax: Proceedings of a Symposium on Iconicity in Syntax, Stanford, June 24-6, 1983, pages 49–72. John Benjamins.
Williamson, Kay (1965). A grammar of the Kolokuma dialect of Ịjọ. Number 2 in West African Language Monographs. Cambridge University Press.

The focus of the present study are serial verb constructions (SVCs) in Bezen, a previously little researched Southern Jukunoid language spoken in one village in North-Western Cameroon. Serial verb constructions are defined as several verbs within a clause that are neither coordinated, nor subordinated. Bezen therfore forms part of a larger linguistic area featuring SVCs reaching from the Ivory Coast to Cameroon (Dimmendaal 2001: 382). Several related Niger-Congo languages have been described to feature SVCs: the Kwa languages Likpe (Ameka 2009), Ewe (Ameka 2006), and Fon (Lefebvre 1991), the Gur language Supyire (Carlson 1994), Benue-Congo Yoruba (Bamgboṣe 1974) and the neighbouring Grassfields-Bantu language Isu (Kießling 2011). The closely related lan-guages Yukuben (Anyanwu 2013) and Kuteb (Koops 1969; 2009) also feature SVCs.
Apart from a description and analysis of SVCs, this work also contains a description of the Bezen verbal system. In the first chapter, previous works on serialisation in related West-African languages are discussed. The data-collection process and the analysis are also presented in this chapter. The verbal system is described in the second chapter, including the aspect and modal categories, adverbs, deictic marking and negation patterns. In this chapter, I also address grammaticalisation processes during which lexemes develop to grammatical morphemes. Since SVCs are defined in delimitation to other multiverb-constructions, such as subordinated and coordinated clauses, these are addressed in the third chapter. The fourth chapter finally is dedicated to the profound discussion of SVCs.
The Bezen data confirms Aikhenvald’s (2006) typology which allows for a classifi-cation of SVCs into two groups, symmetrical and asymmetrical SVCs. Whereas the verbs in symmetrical SVCs are not semantically restricted, at least one verb in asymmetrical SVCs has restricted semantics. Asymmetrical SVCs can be further subclassified: some verbs mod-ify others concerning quality, others introduce an aspectual notion. One group of semanti-cally subordinate verbs introduce arguments, others express the direction of a movement. Switch-function SVCs are a special form of SVCs in which the object of the first verb serves as the subject for the second verb. In Bezen, these SVCs cannot be classified as one of the categories symmetrical or asymmetrical. Some authors (Haspelmath 2015) argue that this type of constructions should not be included in descriptions of serialising languages at all. However, since the verbs share the object of V1, they are included here.

Die vorliegende Arbeit behandelt serielle Verbkonstruktionen (SVKs) in der noch wenig erforschten südjukunoiden Sprache Bezen, die in einem Dorf im Nordwesten Kameruns von nur etwa 450 Menschen gesprochen wird. Als serielle Verbkonstruktionen bezeichnet man mehrere Verben in einem Satz, die weder durch Koordinations­morpheme, noch durch Subordinatoren miteinander verbunden sind. Bezen fügt sich somit in eine lange Reihe verwandter Niger-Kongo-Sprachen ein, die dieses Phänomen aufweisen.