--------------------------------

Celine Romero
University of Toulon
Email: celromero at yahoo dot com

OUGHT TO in Middle English : a semi-grammaticalization?

In Present Day English (hence PDE), OUGHT TO is considered as a semi-modal, something inbetween a lexical verb and a grammaticle item. Yet, that semi-modal, like all the other modals, belonged in Old English (OE) to the same class of verbs : the preterite present verbs.

As shown in Roméro (2005), those verbs all underwent the grammaticalization process by the end of the OE period, along with the causative verbs and the particle TO. However, in Middle English, AGEN still remains apart in that classification since it is mainly followed by TO, which grammaticalized as well. At that period, we could find AGEN as a lexical verb or as a modal verb (either followed by a bare infinitive or the particle TO), as the examples show :

Det hit er ahte. That he already owned it. (CMLAMBX1,313.C2.1087)

Dt euch mon ahte hersumin & herien in eorDe. That each one ought to obey and praise on earth. (CMKATH,23.75)

[Arthure] owen al De worlde TO deme. Arthur ought to deem all the world. (CMPURVEY,I,52.2123)

Then we can question why AGEN followed a different path compared to its OE classmates. In examples displaying TO, it functions as a control verb, and what follows the particle is a CP, whereas the other modals are raising verbs followed by a VP. Could we claim that the grammaticalization of AGEN and the grammaticalization of TO cancelled each other out, like the negative concord process ? Or could we say that, in this particular case, the process of grammaticalization of those two items is not finished yet ? TO came to bear the subjuctive content of the dative-marked infinitive that could follow it in OE, and OUGHT came to be read as a conditional, an irrealis form and not a past form anymore. Moreover, we have found some examples where AGEN was not reanalyzed as a modal, but TO was, and some other examples where AGEN was reanalized but TO was not.

We shall then try and find out (within a Minimalist framework, and with the help of Distributed Morphology) the relations which exist between these two items, in comparison with the other modals, with time and tense, and with grammaticalization to explain this very particular status OUGHT TO has in PDE.

References
The Syntactic Evolution of Modal Verbs in the History of English", PhD Thesis, Paris 3-La Sorbonne Nouvelle, available at http://www.celineromero.com/PhDThesis.html.