Since the first edition of Palmer's 1986 book on mood and modality (and even prior to that), there has been an increasing interest and an increasing number of studies on the expression of modality, both from a typological/cross-linguistic perspective, and in analyses of particular languages. In general, this research has focused on mood, modal verbs, and particles, as these are (cross-linguistically) the most common grammatical systems for expressing modality. However, there are still a number of papers, monographs, etc, (e.g. Nuyts 2001 study on epistemic modality) which analyze lexical expressions of modality (e.g. in epistemic modality the words like probable, probably, possibly, etc).
My motivation for this post comes from several analyses of modal systems (particularly Palmer 2001) which posit, based on empirical analyses of various languages, that languages tend to possess or mainly depend on one grammatical modal system (modal verbs, verbal moods, or particles). For example, when English lost its verbal moods, it began to develop a verbal modal system. German possesses both a fully grammaticalized system of modal verbs, as well as verbal moods, but the use of verbal moods has become increasingly less used in spoken (and even written) german. Many native american languages depend on particles to express modality, and lack both verbal moods and modal verbs.
What interests me, however, is that to some degree classical greek (contrary to the trend revealed by cross-linguistic research) possesses all three systems of modality. Obviously, Greek has a highly developed system of verbal moods. However, in addition to possessing a number of particles, classicgreek has several particles which express various modal aspects. For example, the particle pou is clearly epistemic, and the particle an is laso often used to construe epistemic modality.
Finally, classical greek possesses an (albeit somewhat wealkly developed) system of grammaticalized modal verbs. The best examples are the impersonal uses such as chre/χρῆ or dei/δεῖ. Many of these verbs primarily express deontic modality, but many of the same verbs can express epistemic modality, and some (e.g. eikos/εἰκός) are primarily used to express epistemic modality.
The point (finally) is my confusion/interest in this anamoly in classical greek, and my question is whether anyone is aware of an analysis or analyses of modality in classical greek. I do have Willmott's book on Homeric moods, which deals extensively with modality, but it is not only dealing with a different dialect of ancient greek, it also lacks some fundamental investigations into the greek modal system.