Analogies and blends - new category old ideas

I thought, I might add another category for the new year. analogies. This springs from two separate and seemingly conflicting tendencies in my intellectual life. One is to expose potentially harmful and disparate analogies and the other is to create new analogies to illuminate similarities otherwise unnoticed. I use these strategies in teaching, popular writing and academic research possibly to slight excess but on balance there seems little doubt that, when used with a slight amount of reflective caution, analogies are useful both in their construction and desctruction :"(Stop me before I say 'deconstruction')": .

Now, analogies, in the sense I use them, are nothing more than elaborate metaphors. We could spend years trying to extricate one from the others in this tightly knit weave (spletenec) or family of tropes (irony, metaphor, analogy, etc. and possibly metonymy). We can look at them as either cognitive strategies, textual forms or means of interaction. As cognitive strategies they are a kind of conceptual integration (blending) as described by Fauconnier and Turner - and can be investigated as such. As textual forms, they can but rarely do a purely decorative function (sometimes in the service of so called elegant variation) but mostly they serve to maintain cohesive harmony and coherence of the text. A significant portion of their functionality is to invite some kind of an interaction from the recipient of the text. These three perspectives on analogies are partial and very difficult to separate from one another but they might provide a bit more clarity then the traditional division of these tropes.

These tropes always appear as part of cognitive models (frames, domains, scripts or paradigms). Slightly confusingly, they operate both within models (on submodels) as well as across models but their output is always a new model (a blended conceptual space). Whether we call them metaphors, analogies or by some other name depends on how closely they are integrated, how much is their integration elaborated and what is the nature of their integration. The analogies I will deal with in this category will mostly be explicit, elaborated in some detail, purposeful and project (usually two) complex domain onto one another. They will be fairly loosely integrated in the sense that the identity of the original domains will be preserved and the resulting new model will be a clearly recognizable if slightly modified version of the target domain which the source domain will remain largely (althought not completely) unchanged.
The posts in this category will mostly be random thoughts that come up in my intellectual work rather than discussions of specific texts. They will also always belong in another one or two categories.