Málfræðifélagið

This paper reports the results of a nationwide study of an innovative syntactic construction, exemplified by Það var barið mig or Það var hrint mér, that is developing in the language of young Icelandic speakers. A questionnaire designed to test the syntactic properties of this construction was distributed to 1731 tenth-grade students throughout the country in the academic year 1999–2000.

This number represents 45% of all children born in 1984. The questionnaire was also distributed to 205 adults from various parts of the country. Our results show that the construction is widespread throughout Iceland except in inner Reykjavík, where the acceptance rates were only about half of the acceptance rates elsewhere in the country. Acceptance of the new construction is not correlated with gender, but is strongly correlated with the parents’ level of education. The more highly educated the parents, the less liklely the adolescents are to accept the new construction as part of the language.

In the new construction, the patient argument of a transitive verb like berja ‘hit’ or hrinda ‘push’ remains in object position. If the verb does not govern either dative or genitive, this patient argument is assigned accusative rather than the nominative case that would be found in the canonical passive, and consequently fails to agree with either the finite auxiliary verb or the past participle of the main verb.

Furthermore, this postverbal NP may be definite, in apparent violation of the so-called “definiteness effect” that holds for það-sentences (expletive sentences) generally.

The new construction resembles the accusative-assigning participial constructions which have developed historically from a passive participle in such languages as Irish, Polish, and Ukrainian. In Chapter 5, we outline two possible analyses for the new construction: that it is either (i) a variant of the canonical passive without movement of the object to subject position, or (ii) a syntactically active construction with a phonologically null impersonal subject. We discuss four syntactic properties which distinguish active from passive voice clauses: the occurrence of an agent in a by-phrase, the occurrence of bound anaphors in nonsubject positions, the control of subject-oriented adjuncts and the occurrence of nonagentive verbs. These properties are used to contrast the syntactic behavior of the superficially similar -no/to constructions in Polish and Ukrainian.

This comparison shows that the Polish construction has the syntactic properties of an impersonal active construction, whereas its Ukrainian counterpart behaves like a canonical passive. We then discuss the results of our survey on the innovative construction in Icelandic with respect to the various syntactic properties which bear on its syntactic representation. The results of our survey suggest that despite its apparently passive morphology, the new construction is acquiring the syntactic properties of an active construction. It is noteworthy that the same is true of impersonal passives of intransitive verbs for nearly 50% of adult speakers.

In the last chapter we discuss possible factors that might have motivated this syntactic change in Icelandic.

Umsjón