The regrammation of paradigms: the development of auxiliaries in Danish

Abstract

This article surveys the development of voice and tense auxiliaries in Scandinavian with a focus on Danish. Voice is the first category (documented indirectly by Gothic) to show periphrastic forms in paradigmatic cooperation with inflectional forms; and these periphrastic forms are modelled on predicative constructions. Modern Danish has introduced verbal constructions at the expense of all predicative morphology, and the Old Scandinavian auxiliary verbs hafa/hava ‘have’ and vera/wæra ‘be’, verða/wartha ‘become’ have undergone semantic changes, including specialisation and markedness shift. This regrammation process involves semantic changes that must be described in terms of paradigm (re)organisation and cannot be captured in terms of changes along the parameters of the cline of grammaticalisation. Two main types of auxiliaries are distinguished: inflectional auxiliaries, with predication scope (tense, mood and voice) as part of otherwise inflectional paradigms, and, constructional auxiliaries (copula verbs) with predicate scope. Modern Danish inflectional auxiliaries express perfect tense and active voice with transitive verbs (have ‘have’), perfect tense, active voice and telicity with intransitive verbs (være ‘be’ and have ‘have’), stativity (copula verbs være ‘be’ and blive ‘become, turn’) and passive voice and telicity in the periphrastic passive (være ‘be’ and blive ‘become, be’). Their meaning potentials differ according to the paradigms they are part of.

Keywords:

• Auxiliaries
• Danish
• paradigmatic organisation
• regrammation

Notes

¹ Haspelmath (2002, 143, 144) insists on the Latin gap-filling model and regards the privative model as pure syntax irrelevant for morphology.

² Substantially, this view places Anderson on a par with at least some generative analyses, Faarlund (2004, 127) being a clear example: The auxiliary (for instance, the copula of the progressive forms) is semantically empty, but can nevertheless unidirectionally govern something called its ‘complement’.

³ Eisenberg (2006, 35) offers the relation between noun and genitive as a simple example of endocentrism.

⁴ The Gothic examples are from Streitberg (1965). L = Luke, J = John, R = Romans, Sk = Skeireins (Commentary to St. John). Examples: L 14, 11 = Luke 14, verse 11; Skeireins III, 9 = Skeireins fragment III, line 9.

⁵ Johannisson (1945) recapitulates the classical discussion and the state of affairs: Neither view can stand alone, and the bleached construction (22) is the construction to be reinterpreted, but the supine must come from somewhere else (24bc). (24b) is a special case of the direct object construction.

⁶ Faarlund (2004, 128) seems to count all these verbs as auxiliaries, but they cannot be, neither by his definition nor by mine, since they assign case roles to their subjects, or in my argot: govern their subjects lexically.

⁷ Larsson (2015, 163, 164) refers to SAG, for examples, like dom är redan anlända ‘they have already arrived’ and Frida är just hitflyttad från Stockholm ‘Frida has just moved to here from Stockholm’. Both anlända and hitflyttad lexicalise telicity and unlike Danish, the addition of telic adverbs and PPs is not enough to open up for vara: *Han är sprungen till affären ‘he has run to the store’ (2015, 172).