Stellenbosch Papers in Linguistics Plus, Vol. 65, 2022, 1-9
doi: 10.5842/65-1-965
The rise and fall of Serial Verb Constructions: Preamble
Alexander Andrason
Faculty of Humanities, University of Cape Town, South Africa
E-mail: aleksand@hi.is
Alexandra Y. Aikhenvald
Jawun Research Centre, Central Queensland University, Australia
E-mail: a.aikhenvald@cqu.edu.au
Abstract
This is a brief introduction to the special issue of Stellenbosch Papers in Linguistics Plus. We
present the concept of serial verb constructions (SVCs) conventionally understood as
monoclausal sequences of verbs without any overt marker of coordination, subordination, or
syntactic dependency. We then focus on the mechanisms at work in the evolution of serial verb
constructions, and the investigations of their origin and demise. We introduce the prototype
approach to the category of SVCs as the basis of the study of verb serialization throughout the
volume and discuss the research strategies applicable to the development of serial verbs in
individual languages. The concluding section offers an overview of the volume.
Keywords: Serial verb constructions; diachrony; prototype; dynamization of typology
1.
Setting the scene
The diachrony of Serial Verb Constructions (SVCs) – and thus the shape and the properties of
the developmental paths linking the origin of these types of grammatical structures to their
demise or to their evolution and expansion – is one of the research areas within the scholarship
of verbal serialization that is in urgent need of a broad range of systematic and empirical
studies.
Indeed, in her comprehensive 2018 monograph, Alexandra Aikhenvald states that
“determining the origins of serial verbs is a strenuous task [... since for] no language family in
the world do we have enough historical evidence to confidently trace the roots and the
development of [these types of] constructions” (Aikhenvald 2021:196). This is particularly true
of many African, American, Asian, and Oceanic languages which, on the one hand, tend to
abound in serializing patterns, but on the other hand, usually lack written records that would
allow us to trace the developmental pathways of SVCs. However, even for languages whose
historical evolution is well documented in writing, sometimes going back in time for several
© 2022 The authors. This work is licensed under the Creative Commons Attribution 4.0 International License.
2
Andrason & Aikhenvald
centuries and even millennia – as is true of many Indo-European and Semitic languages – no
systematic diachronic studies on SVCs have been carried out thus far. 1
Serial verb constructions, or serial verbs, are conventionally understood as sequences of verbs
without any overt marker of coordination, subordination, or syntactic dependency of any sort.
Serial verbs are monoclausal constructions describing what is conceptualized as a single event.
They share prosodic properties with monoverbal constructions. A serial verb has one tense,
aspect, mood, modality, and evidentiality value — that is, one component cannot refer to past
and another to present. The components of a serial verb cannot be negated or questioned
separately from the whole construction. Each component must be able to occur on its own (a
brief survey and a bibliography of serial verb constructions is in Aikhenvald 2018). We return
to their further properties and their classification into asymmetrical, symmetrical, and event
argument types in Section 2.
Despite the abovementioned dearth of direct diachronic evidence and a similar scarcity of
studies dedicated specifically to the evolution of serializing patterns, a number of
developmental generalizations have been posited, mostly through comparative methods and
synchronic typological research. The first class of these tendencies concerns the origin of
SVCs and their gradual development towards fully fledged serializing structures (Aikhenvald
2006; 2018; 2021):
(a)
(b)
(c)
(d)
(e)
(f)
SVCs result either from (i) clause fusion, (ii) verbal modification, or (iii)
concurrent grammaticalization.
While the first two developmental scenarios (i.e., clause fusion and verbal
modification) account for the development of both symmetrical and asymmetrical
SVCs (see section 2 below), the third scenario (i.e., concurrent
grammaticalization) is limited to the emergence of asymmetrical SVCs.
Asymmetrical SVCs that express direction/orientation, aspect, extent, and change
of state develop faster than modal, valency-increasing, and argument-adding
SVCs, with SVCs used for comparative, valency-decreasing, and other purposes
developing last.
The emergence of two-component SVCs precedes the development of more
elaborated structures involving three or a larger number of verbs.
Symmetrical SVCs evolve only once asymmetrical SVCs have been developed.
In other words, no language has symmetrical SVCs unless it also has
asymmetrical ones.
The emergence of SVCs compensates for the reductive processes affecting
inflectional morphology, thus being correlated with the expansion of analytical
strategies.
1
This scarcity of diachronic research on SVCs in the Indo-European and Semitic families may stem from the fact
that verbal serialization has generally been associated with the languages of Africa, the Americas, Southeast Asia,
and Oceania. In contrast, serial or similar constructions found in Indo-European and Semitic languages have
traditionally been classified as different grammatical categories (e.g., hendiadys, auxiliary constructions,
Koppelung, pseudo-coordination, coordination, and several others). Only recently, mostly due to the adoption of
a prototype-driven approach to verbal serialization, which is a more flexible and gradient and essentialist
definition (see section 2 below), these same structures started to be analyzed as subtypes of SVCs.
http://spilplus.journals.ac.za
The rise and fall of Serial Verb Constructions
3
Evidence related to the development of SVCs after reaching the stage of a fully-fledged SVC,
including their demise or dissolution into other grammatical categories, is slightly more
abundant although the generalizations proposed thus far also rely heavily on comparative and
typological research rather than direct diachronic data. The following tendencies have been
postulated (Aikhenvald 2006; 2018; 2021):
(a)
(b)
(c)
(d)
The endpoint of the development of asymmetrical SVCs is different from that of
symmetrical SVCs: asymmetrical SVCs undergo grammaticalization, while
symmetrical SVCs undergo lexicalization.
Functionally, minor verbs in asymmetrical SVCs tend to evolve into markers of
tense, aspect, and modality, including evidentiality, as well as expressions of
directional, locative, comparative, and superlative domains.
During that evolution, minor verbs in asymmetrical SVCs are decategorized: they
evolve into auxiliaries, adpositions, particles, conjunctions, complementizers,
and a range of bound morphemes, eventually losing their verbal status entirely.
Symmetrical SVCs develop into lexical units. The two (or more) verbs merge into
a single verb, ultimately contributing to the expansion of the verbal lexicon of a
language.
Importantly, both during the development towards an SVC and the development from an SVC
towards other grammatical categories, the changes, including those specified above, are
gradual. This graduality gives, in turn, rise to fuzzy cases characterized by categorial ambiguity
and/or forms that allow the association with far more than one grammatical category.
Furthermore, although the development of SVCs may be language-specific and constitute a
language-internal feature, it may also be genetically motivated (i.e., typical of a family of
related languages) and/or due to language contact and areal diffusion.
The present issue of Stellenbosch Papers in Linguistics PLUS is the first collection of articles
dedicated specifically to the evolution of SVCs and their dynamics, in particular the rise and
fall of these types of constructions. The exact and narrow aim of this volume is to test the
above-mentioned generalizations on a large and phylogenetically, typologically, and
geographically diversified language sample. In that manner, we aspire to provide a new body
of evidence that could shed more light on how SVCs are evolutionarily and conceptually
related to other categories, both those from which they derive and those towards which they
evolve.
2.
Serial Verb Constructions: the prototype approach
The category of SVCs constitutes an extremely complex phenomenon that encompasses a wide
range of diverse constructions. In our view, a prototype approach to linguistic categorization
provides the most suitable model that allows one to both preserve the diversity of SVCs attested
across the languages of the world and ensure the conceptual unity of the SVC category and its
status as an autonomous grammatical taxon.
In accordance with a prototype approach to linguistic categorization (see Evans and Green
2006; Janda 2015), the prototype of an SVC is defined cumulatively as a set of features. The
inclusion of a feature into the prototype is motivated by at least one of the two reasons. On the
one hand, some prototypical features are typologically pervasive, i.e., commonly attested in
SVCs found across the languages of the world. On the other hand, some prototypical features
are cognitively salient and distinguish SVCs from other grammatical categories most
http://spilplus.journals.ac.za
4
Andrason & Aikhenvald
efficiently. Arguably, all the features included in the prototype tend to converge diachronically.
That is, the prototype acts as an evolutionary attractor: SVCs develop towards a state in which
all the prototypical properties are instantiated.
The following features have been considered as prototypical and jointly define the prototype
of an SVC (Aikhenvald 2006; 2011; 2018; 2021; Dixon 2006):
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
A prototypical SVC makes use of (at least) two verbs.
Each component of a prototypical SVC can be used on their own outside an SVC
pattern; that the verbal components may function as full verbs (i.e., semantically
robust, referential content lexemes) and predicates of independent clauses.
The components of prototypical SVCs are not connected through clause
combining markers. This thus excludes the presence of coordinators,
subordinators, complementizers, or any other markers of syntactic dependency.
A prototypical SVC exhibits a single value for tense, aspect, and mood/modality,
as well as evidentiality, polarity, and illocutionary force.
The components of prototypical SVCs are marked by the same (or nonconflicting) tense-aspect-mood-modality-evidentiality markers.
A prototypical SVC exhibits a unitary argument structure. Crucially, the verbal
components share their subject referent. Separate subject arguments are thus
disallowed.
In a prototypical SVC, operators of time, place, and manner/instrument/means
and similar operate jointly over the verbal components. Conversely, the scope of
these operators is not limited to one verb only.
In a prototypical SVC, subordinating markers and nominalisers are shared by all
the verbal components. This means that the use of participial, gerund, and
infinitival forms is not limited to one verb only and has scope over the whole
construction.
In a prototypical SVC, the verbal components are not separated prosodically by
means of contouring, comma intonation, pause, or any type of bi-phrasal/clausal
phrasing.
The above properties imply that a prototypical SVC attest to mono-eventhood
(i.e., it expresses a single event) as well as mono-clausality and monopredicativity (i.e., it belongs to a single clause in which it forms a single
predicate). 2
Some properties enumerated above attest to the constructional cohesiveness of an SVC
prototype (see (c)-(i)), while a few others reflect the individuality of its formative components
and thus somewhat lesser constructional cohesiveness than is the case of, for instance,
monoverbal predicates and synthetic tenses, aspect, and moods, modalities, and evidentiality
forms (see (a)-(b)).
The prototype is a critical element in the category: it constitutes its conceptual nucleus and a
point in relation to which the categorial status of all the other members is measured. Members
that instantiate the prototype fully, complying with all or nearly all prototypical features are
viewed as canonical. In contrast, members that instantiate the prototype minimally, complying
with only a few prototypical features are non-canonical. Between these two extremes, there is
2
For a discussion of the several violations of all the prototypical features introduced in this section, consult
Aikhenvald (2006; 2018; 2021).
http://spilplus.journals.ac.za
The rise and fall of Serial Verb Constructions
5
a cloud of semi-canonical members that instantiate the prototype partially, complying with
several prototypical features, although neither all nor extremely few. Overall, the prototype
structures the SVC category and provides it with a topological model (i.e., spatial
visualization). That is, by correlating the extent of compliance of each member with its position
relative to the prototype – the more canonical the closer it is to the prototype and, inversely,
the more non-canonical the more remote from it, it is – the category adopts the form of a radial
network (on radial networks see Evans and Green 2006; Janda 2015).
The canonical, semi-canonical, and non-canonical profiles exhibited by the respective
members of an SVC category can be interpreted in dynamic terms as attesting to different
evolutionary stages available to SVCs. In general, the more non-canonical an SVC is the more
evolutionarily distant it is from the prototype. This distance itself may correspond to and stem
from two phenomena. Some cases of lesser canonicity emerge because of an incomplete
advancement along the path travelled by SVCs and those constructions’ insufficient
cohesiveness. As these SVCs appear in stages that precede the stage of the prototype, they are
referred to as pre-canonical. However, lesser canonicity may also correspond to and stem from
an excessive advancement along the path travelled by SVCs and those constructions’ excessive
cohesiveness. These SVCs appear in stages that follow the stage of the prototype and are
therefore referred to as post-canonical. The least canonical members, whether of a precanonical or post-canonical character, are related to other grammatical taxa; namely, those that
are diachronically prior (i.e., from which insufficiently cohesive non-canonical SVCs have
derived) and posterior (i.e., towards which excessively cohesive non-canonical SVCs are
subsequently developing (Andrason 2019; Andrason & Koo 2020; see also Andrason 2016;
Georgakopoulos & Polis 2018).
Among the various types of SVCs, two main classes are distinguished: symmetrical SVCs and
asymmetrical SVCs. In an asymmetrical type, the verbal components used are semantically
uneven. One verb, referred to as a ‘major’ verb, specifies the type of action or activity that is
expressed by the entire construction. This verb constitutes the variable element in the
construction and draws from a non-restricted (open) class. The other verb modifies the event
expressed by the ‘major’ verb in terms of aspect, mood/modality, direction/orientation,
valency-increasing terms, etc. This verb, referred to as a ‘minor’ verb, draws from a
semantically restricted (closed) class and constitutes the constant element in the construction.
In contrast, in a symmetrical type, both verbs contribute relatively equally to the constructional
behavior of an SVC in the way that “none of them determines the semantic or syntactic
properties of the construction as a whole” (Aikhenvald 2006:22). The verbs used in
symmetrical SVCs do not draw from semantically and grammatically restricted classes and
typically denote a sequence of sub-actions, cause-effect, or manner in which the action is
performed. As mentioned in Section 1, both types tend to follow different evolutionary
scenarios. Asymmetrical SVCs undergo grammaticalization, whereby the minor verb acquires
and/or increases its grammatical function. In contrast, symmetrical SVCs undergo
lexicalization, whereby the verbs used develop towards a single lexeme or idiom (Aikhenvald
2006; 2018; 2021).
http://spilplus.journals.ac.za
6
Andrason & Aikhenvald
3.
Research strategies
There are two main strategies with which one can study the evolution of SVCs and propose or
verify developmental generalizations pertaining to these types of constructions. One strategy
is purely diachronic, while the other draws, principally, on the examination of synchrony.
The first, diachronic, strategy involves the study of SVCs at different historical periods within
the same language. This type of research is the scarcest in SVC scholarship because, as we
explained in the introductory section, most languages have not been documented in written
texts throughout their history. However, diachronic research is probably the most valuable: it
provides direct empirical evidence that reveals how SVCs have changed over more or less
extensive time, enabling us to eventually trace the entire grammatical life of these constructions
from their origin to death. This strategy is therefore the best technique to test the evolutionary
generalizations that have been proposed in any other manner (see the next paragraph), perhaps
the only one that can ultimately corroborate, falsify, or nuance them.
The other strategy – or rather cluster of strategies – draws on the method that may be referred
to as dynamization of synchrony (cf. Jakobson 1962:650-652, Croft 2003). With this strategy,
a synchronic state exhibited by an SVC in a specific language at a specific point in time or a
collection of such synchronic states exhibited in several languages, are interpreted
diachronically and arranged into a sequence. This sequence, in turn, reveals and/or reflects an
evolutionary path that these constructions have followed. As mentioned above, the more noncanonical the profile of an SVC is, the more remote from the stage occupied by the prototype
and thus any canonical instantiation. If this non-compliance is explained in terms of an
insufficiently cohesive SVC profile, the progression towards the SVC prototype is incomplete
and that SVC occupies a stage that must precede the stage of the prototype. If non-compliance
stems from excessive cohesiveness, that SVC has advanced beyond the stage of the SVC
prototype and thus occupies one of the stages that follow the stage of the prototype.
This dynamization of synchrony itself involves three related, yet slightly distinct strategies:
(a)
(b)
(c)
Dynamization of states exhibited by closely related languages that have been
spoken at different periods of time. Such varieties may belong to a linguistic
branch or larger family. This strategy may, in some way, involve a diachronic
perspective – at least from the family’s perspective – as the varieties compared
can document (a) evolutionary trend(s) shared by the members of a certain
phylogenetic group across centuries and even millennia.
Dynamization of synchronic states exhibited by contemporaneous languages and
thus temporarily concurrent. Such varieties may be closely related genetically (as
in the type explained above), very remotely related, or entirely unrelated.
Dynamization of a synchronic state exhibited by a single language. In this
approach, the state(s) exhibited by an SVC in a specific variety at a specific time
is (are) depicted as a semantic map, which in turn can be interpreted
diachronically, i.e., matching a certain fragment of the evolutionary path along
which this SVC travels (Andrason 2016).
http://spilplus.journals.ac.za
The rise and fall of Serial Verb Constructions
4.
7
Structure of this volume
The organization of the present volume reflects the various strategies enabling one to study the
evolution of SVCs, which we introduced in the previous section.
We begin our volume with three articles that provide direct diachronic evidence (see Part 1).
Each of these studies demonstrates the development of SVCs in a specific language across a
determined period of time. Unsurprisingly, all these languages belong to the Indo-European
family – whose history is, as explained previously, extensively documented in written records
– specifically, its Balto-Slavic branch. Alexander Andrason, Małgorzata Gębka-Wolak and
Andrzej Moroz analyze the history of some SVCs in Polish, Daniel Weiss conducts a similar
study on Russian, and Nicole Nau on Latgalian. These three contributions concentrate on
constructions which involve the verb ‘take’ in the respective languages and express unexpected
and sudden action. 3 The authors consider both language-internal and language-external
motivations for the creation and expansion of these constructions.
The remaining part of the volume encompasses articles that exploit dynamization of synchrony
as their main research strategy. The first group of these works dynamizes the synchrony of
related languages that have been spoken at different historical epochs (see Part 2). As the full
extent of the time period covered by these languages ascends to one thousand years, this type
of dynamization also has a diachronic dimension. The languages studied in this way belong to
the North-West branch of the Semitic family – another linguistic family with extensive written
records. Alexander Andrason examines the development of verbal serialization in CanaanoAkkadian, Ugaritic, Biblical Hebrew, and Biblical Aramaic. Christian Locatell examines the
state of SVCs in Old Aramaic.
The second group of articles making use of dynamization of synchrony exploits purely
synchronic states of languages that are attested currently (see Part 3). Daniel Ross and Joel
Lovestrand draw their evolutionary generalizations from a genetically, typologically, and
geographically diversified sample of 325 languages. Two further articles formulate
developmental proposals given the states exhibited by SVCs in closely related languages.
Anne-Maria Fehn and Admire Phiri analyze several varieties from Northeastern Kalahari
Khoe: Ts’ixa, Danisi, Gǁoro, Shua, Deti, Tjwao, TcireTcire, and Gǁabak’e. Lee Pratchett
studies Juǀ’hoan and ǃXun from the Ju language complex of the Kx’a family.
The third group of papers focuses on the dynamization of the synchronic states exhibited by
SVCs in a single language, and the development of newly emergent serial verbs (see Part 4).
Aikhenvald examines SVCs in an Arawak variety, Tariana, a North Arawak language, while
Ronald Schaefer and Francis Egbokhare offer an in-depth analysis of Emai, an Edoid language.
At the end of this volume, we and the other authors jointly discuss our main findings and their
implications for SVC scholarship. We set out the main points of agreements as well as certain
issues where agreement among us is more difficult to reach, thus suggesting possible lines of
future research.
3
Similar double verb constructions with the first component ‘take’ have been attested in Portuguese and Spanish
(see the discussion Aikhenvald 2021:125, 140, and Coseriu 1966), and also some Finno-Ugric and Turkic
languages (Pukkinen 1966; Csató 2001).
http://spilplus.journals.ac.za
8
Andrason & Aikhenvald
References
Aikhenvald, A. Y. 2006. Serial verb constructions in typological perspective. In A. Aikhenvald
and R. M. W. Dixon (eds.) Serial Verb Constructions: A Cross-linguistic Typology. Oxford:
Oxford University Press. pp. 1-68. https://doi.org/10.1093/oso/9780198791263.003.0010
Aikhenvald, A. Y. 2011. Multi-verb constructions: Setting the scene. In A. Aikhenvald and P.
Muysken (eds.) Multi-verb Constructions: A View from the Americas. Leiden: Brill. pp. 126. https://doi.org/10.1163/ej.9789004194526.i-313.8
Aikhenvald, A. Y. 2018. Serial verb constructions. In M. Aronoff (ed.) Oxford Bibliography
Online. New York: Oxford University Press.
Aikhenvald, A. Y. 2021. Serial verbs. Paperback edition with revisions. Oxford: Oxford
University Press.
Andrason, A. 2016. From vectors to waves and streams: An alternative approach to semantic
maps. Stellenbosch Papers in Linguistics 45: 1-29. https://doi.org/10.5774/45-0-211
Andrason, A. 2019. A pseudo-coordinated Serial Verb Construction “wziąć i V2” in Polish.
Slovo a Slovesnost 80: 163-191.
Andrason, A. and B. Koo. 2020. Verbal serialization in Biblical Aramaic – A dynamic network
approach. Altorientalische Forschungen 47(1): 3-33. https://doi.org/10.1515/aofo-2020-0001
Coseriu, E. 1966. ‘Tomo y me voy’: ein Problem vergleichender europäischer Syntax. Vox
romanica 25(1): 13-55.
Croft, W. 2003. Typology and Universals. Cambridge: Cambridge University Press.
https://doi.org/10.1017/cbo9780511840579
Csató, E. 2001. Turkic double verbs in a typological perspective. In K. H. Ebert and F. Zúñiga
(eds.) Aktionsart and Aspectotemporality in Non-European Languages. Zürich: Universität
Zurich. pp. 175-187.
Dixon, R. M. W. 2006. Serial verb constructions: Conspectus and coda. In A.Y. Aikhenvald
and R. M. W. Dixon (eds.) Serial Verb Constructions: A Cross-linguistic Typology. Oxford:
Oxford University Press. pp. 338-350. https://doi.org/10.1515/lingty-2017-0008
Evans, V. and M. Green. 2006. Cognitive Linguistics: An Introduction. Edinburgh: Edinburgh
University Press.
Georgakopoulos, T and S. Polis. 2018. The semantic map model: State of the art and future
avenues for linguistic research. Language and Linguistic Compass 12(2): 1-33.
https://doi.org/10.1111/lnc3.12270
Jakobson, R. 1962. Retrospect. In R. Jakobson Selected Writings. Vol I. Phonological Studies.
Hague: Mouton & Co. pp. 629-658.
http://spilplus.journals.ac.za
The rise and fall of Serial Verb Constructions
9
Janda, L. 2015. Cognitive Linguistics in the Year 2015. Cognitive Semantics 1: 131-154.
https://doi.org/10.1163/23526416-00101005
Pulkkinen, P. 1966. Asyndeettinen rinnastus suomen kielessä [Asyndetic Coordination in
Finnish]. PhD dissertation, University of Helsinki.
http://spilplus.journals.ac.za