Clearly the semantics of the pre-modals underwent significant change along the trek toward full auxiliarihood. The meanings of the individual words themselves evolved, as did their relationships with other pre-modals. For a detailed account of the interplay of the various senses of the pre-modals throughout their history, see Traugott (1972a). What follows is a brief synopsis.
The original "notional" meanings of the pre-modals have already been mentioned,
but I will repeat them for convenience: cann 'have the intellectual capability, know, know
how to'; m�g 'have the physical ability, be able to'; mot 'be allowed, may'; sceal 'owe,
ought'; wile 'intend, wish, desire'. These were the primary meanings of the pre-modals
during OE, which included a range of root and epistemic modal uses. Epistemic modality
refers to the subject's attitude about the truth of the sentence, while root readings involve
notions of permission and obligation (deontic), and possibility and necessity (dynamic).
Futurity is also sometimes considered a separate type of modal meaning. Permission was
most often expressed by mot, obligation by sceal, volition by wile. During the course of
ME, may began to displace mot for expressing permission, and was prevalent from the
fourteenth century. Meanwhile, can was extended to signify ability in general, both
intellectual and physical. Must then came to be used to express obligation, as shall
gradually lost that interpretation. In ME and eModE, shall and will both came to be
commonly used to express prediction, though as early as late OE, both were being used
almost as future tense markers. Only in the 1700s did can come to express permission.
Shepherd (1982) theorized that modality in general develops from root (deontic) to
epistemic, and indeed, root readings are common for OE modals, continuing to thrive into
ME (90-96) (Denison 1993: 300-303).
In OE, m�g, sceal, and wile, when used personally, show the first traces of
epistemic use, and in ME examples can also be found, though they are quite scarce (97-
100). Frequent early examples of epistemic modality can, however, be found in
impersonal constructions (Denison 1990, 1993; Warner 1993), which died out in general
during ME (101-112).
More interesting than this, however, was the erosion of the tense relationship between the present and past forms in such pairs as cann/cu�e, m�g/meahte, wile/wolde, sceal/sceolde, and mot/moste, where the only semblance left of a tense relationship is in reported speech. This breakdown in relations began even in OE, and has been attributed to the loss of a distinction between the weak preterite indicative and subjunctive (Goossens, p.149). Sceolde, for example, could be interpreted as either indicative 'had to' or a timeless subjunctive form (113, from Warner 1993, p.149, imposes present obligation). As a result, the members of the pairs came to be viewed as separate lexical items, developing independent meanings.
Some pre-modals are themselves thought to have split into separate lexical items. Warner (1982) suggests this for shall (futurity vs. obligation). Plank mentions the separate verb con/cun, which is based on pre-modal can, and could, which retained its etymological spelling couth/coud(e) when used with the original (non-modal) meaning 'know'.
Next Section
Contents