The Ewe sentence translating literally as 'She take knife cut meat' contains two verbs with no conjunction between them, shares a single tense marking, and the knife is understood as both what she took and what she used to cut. What syntactic property most clearly identifies this as a serial verb construction rather than a coordinate clause?
AThe absence of a subject pronoun before the second verb
BThe fact that both verbs are in the same tense
CArgument sharing — 'knife' serves as the object of 'take' and simultaneously as an instrument for 'cut,' with no overt connective linking the predicates
DThe construction describes a single unitary event rather than a sequence
Argument sharing is the key diagnostic for SVCs. In true SVCs, an argument of the first verb functions simultaneously as an argument of the second verb — here, 'knife' is both the object taken and the instrument used. This co-indexing of arguments across verbs, with no conjunction or subordinator, is structurally different from coordination (where each clause has its own argument structure) and from purposive complements (which have their own subject). Shared tense (option B) is a feature of SVCs but not uniquely diagnostic; many coordination constructions share tense. The absence of subject repetition (option A) is common but not the defining property.
Question 2 Multiple Choice
Why do many European-language speakers initially misanalyze serial verb constructions as grammatical errors or as coordinate clauses with missing conjunctions?
ABecause SVCs only occur in unwritten languages, making them unfamiliar to literate speakers
BBecause the one-verb-per-clause tendency is so deeply entrenched in European languages that multi-verb sequences feel like violations rather than legitimate structures
CBecause European linguists have not studied African and Asian languages carefully enough to recognize the pattern
DBecause SVCs are derived from historical processes of grammaticalization that don't occur in European languages
European languages — especially English, French, German — enforce a strong one-main-verb-per-finite-clause pattern. When European-language speakers encounter a sentence with two verbs and no conjunction, their grammatical intuitions flag it as incomplete or erroneous, because in their native language such a sequence would be ungrammatical. This is a typological bias, not a universal of grammar. SVCs reveal that the one-verb-per-clause property is a feature of one set of languages, not a necessary property of human language. Recognizing this requires the typological perspective that looks for cross-linguistic patterns rather than measuring all languages against a European template.
Question 3 True / False
Serial verb constructions are found extensively in West African languages (Ewe, Yoruba, Akan) and many East Asian languages (Mandarin, Vietnamese, Cantonese), demonstrating that the one-verb-per-clause pattern in European languages is a typological tendency, not a universal of grammar.
TTrue
FFalse
Answer: True
SVCs are robustly attested across unrelated language families — Kwa (West Africa), Niger-Congo more broadly, and Sino-Tibetan, Tai-Kadai, and Austroasiatic families in East and Southeast Asia. Their wide distribution in genealogically unrelated languages suggests they are a genuine typological category, not an accident of borrowing or common ancestry. The fact that they appear so consistently in languages that have no historical connection to each other confirms that the one-verb-per-clause pattern is not a universal constraint on human language but a feature that some language families have grammaticalized and others have not.
Question 4 True / False
Serial verb constructions are best understood as coordinate clauses in which the conjunction ('and' or 'then') has been deleted or omitted in casual speech.
TTrue
FFalse
Answer: False
SVCs are not reduced coordination — they have distinct structural properties that set them apart. In coordination, each clause maintains its own argument structure; arguments are not shared between clauses except through overt pronouns or ellipsis. In SVCs, the shared argument is a defining property, not a reduced element. Additionally, SVCs have unified tense/aspect marking: a single tense marker scopes over the whole construction, whereas coordinated clauses can each take independent tense markings. The argument-sharing and unified tense properties are not derivable from coordination-with-deletion; they require a distinct structural analysis. SVCs are a legitimate grammatical category, not impoverished coordination.
Question 5 Short Answer
What is the significance of 'argument sharing' as a diagnostic for serial verb constructions, and why does it distinguish SVCs from both coordination and subordination?
Think about your answer, then reveal below.
Model answer: Argument sharing means that a single NP functions as an argument of two or more verbs simultaneously — typically, the object of the first verb serves as subject or instrument of the second, with no overt pronoun, conjunction, or subordinator connecting them. This is structurally impossible in coordination: two coordinated clauses each have their own argument structure, and any shared participant is referenced by a repeated noun or pronoun in the second clause. In subordination, the embedded clause has its own predicate and argument structure, typically marked by a complementizer or infinitive. The SVC's shared argument with no connective indicates that the two verbs form a single complex predicate — they jointly describe one event structure with shared participants, not two events linked by logical or temporal connectives. This property forces syntacticians to reconsider what it means for a clause to be a clause.
The diagnostic significance goes beyond taxonomy: argument sharing reveals that in SVC languages, the verb phrase can be genuinely complex — hosting multiple verbal heads that share syntactic positions — in ways that European-language grammars don't permit. This has implications for Universal Grammar: if some languages allow multi-headed VPs and others don't, the variation itself is a datum about the range of possible human grammars.