When we know a language well we are likely to recognise (if only intuitively) that certain collocational associations are acceptable while others are not, yet we may not appreciate that the suitability of such partnerships depends, even now, upon where the words came from in the first place. When words move out of their original domains or habitats they often do so as metaphors or metonymies, and since at the heart of every metaphor lies a metonym, which will be related to fellow inhabitants of the original domain, this inter-relationship will continue to limit each term's freedom of association. Such terms are only likely to co-occur with fellow members of their original habitat, both presupposed and entailed, of which they are the representative metonymic agent. This was so when the terms were first coined and remains so even after usage may have led their surface meanings to undergo changes. The writer looks at four word-sets of ostensible synonyms which demonstrate the argument.
As Sapir (1921) and Haiman (1995) note, grammaticalization of emotional concepts is much less common than that of concepts like modalities. In this paper, I will investigate so-called “emotional should, ” and argue that it is one of the rare instances of grammatical forms encoding emotional concepts. In the literature, it is often analyzed to express non-factual concepts, as other uses of quasisubjunctive should do, in the effort to treat all uses of quasisubjunctive should in exactly the same way. This enterprise, however, runs into serious difficulties. Of course, I am also conscious of the ontogenetical relationship between emotional should and other uses of quasi-subjunctive should, but in the former, the main function has shifted to signaling a speech act which I call “emotional expression.” As for the question of how this development occurred, it will be proposed that emotional should was derived from epistemic uses of quasi-subjunctive should through “discourse-strategic usage extension, ”i.e.extension of the usage of a grammatical form which aims to linguistically express a new speech act utilizing its pragmatic effect. As such, emotional should can be conceived of as a regular outgrowth of quasi-subjunctive should.
There is a good deal of discussion about the relevance of verbal morphology to V-to-I movement. The assumption that rich inflection triggers overt verb movement to Infl may be called the Rich Agreement Hypothesis (RAH) (cf. Bobaljik 2001). The diachronic corollary of the RAH is that the loss of rich inflection leads to the loss of V-to-I movement, and some specific proposals are made as to exactly how to define “rich” inflection. However, few serious att e mpts have so far been made to apply this hypothesis to Verb Second (V2) phenomena. A major reason for this is that apparently rich inflection is not a necessary condition for V2 movement; thus, Mainland Scandinavian languages (i.e., Danish, Swedish and Norwegian) exhibit overt V2 movement in main clauses despite their flat verbal morphology. The fact about Mainl a nd Scandinavians at first glance seems to constitute a crucial evidence against the RAH on V2 movement. This paper will nevertheless show that the loss of V2 in English is closely linked to the decline of verbal inflection from Middle English (ME) to early Modern English and thus can be explained by the RAH. Consequently, I will suggest that there are two types of V2 that occur in different components of the grammar, that is, PF and the syntax.