Base (in Cognitive Grammar)

In Cognitive Grammar semantic structures generally are expected to consist in a designatum which is profiled against a background of related cognitive structures. These related cognitive structures collectively constitute the base of the particular semantic structure.

This distinction of profile vs. base is taken to be a manifestation of figure/ground organization: the profiled element (designatum) is the figure and the base is the ground against which it is construed.

Comments
Given the assumption of encyclopedic meaning, everything that is conventionally associated with the designatum is, at least potentially and in its degree, part of the base. Thus for a word like book, the semantic structure BOOK will designate (profile) the Thing we know as a book against an extremely complex background including knowledge of reading and writing, printing, bookmaking, authoring, libraries, bookstores, bookshelves, indexes and tables of contents, Bibles and Qur'ans, textbooks and cookbooks, etc. virtually ad infinitum.

This might seem at first to lead to massive problems of duplication of meanings and inclusion paradoxes. Thus AUTHOR and BOOKSELLER are included in the meaning BOOK, and BOOK in each of them, and BOOKSTORE and LIBRARY in each of them and vice versa. This paradoxical duplication is the same as was classically recognized for cases like PARENT and CHILD, each of which is needed to define the other, so this is not a new problem. It is recognized in Cognitive Grammar as a pseudo-issue engendered by the container metaphor, our thinking of lexical structures as “containers” for meaning, coupled with our knowledge that physical containers cannot interpenetrate each other. If one conceives instead of a vast network of conventionally established and multiply-related concepts, one can see a lexical item as opening a window on or point of access for a portion of that network. Words like BOOK, AUTHOR, PUBLISH, LIBRARY, and so forth, open windows on the same corner of the network, and naturally the structures in that corner, albeit with different perspectives and different prominences, are accessible from the different windows. There is no duplication because it is in fact the same structures that are accessed. Similarly, PARENT and CHILD both access the same base structure of the generational relationship: one profiles one participant in that relationship and the other profiles the other participant. The designatum of CHILD is part of the base of PARENT, and vice versa, but it is not designated when viewed from that window.

Often one can describe different parts of the base as pertaining to different cognitive domains. Within any domain some specifications will usually be more central (i.e. more relevant, more likely to be activated on any particular occasion of use) than others. And some domains will be more central than others.

For instance, for the English word uncle, the domain of kinship is primary, while domains such as 3-dimensional space, socially-approved attitudes towards people, or age, are clearly less central. Within the kinship domain, the relationship between the male person profiled (designated) and the child of a sibling is clearly the most central specification. The sibling's child (the “ego”) and his or her relationship to the profiled person are so central as to be activated on virtually every occasion of use of the concept UNCLE (thus part of what other theories would call the "denotation"), but they are not profiled (not part of the designatum), but are rather part of the base, the background against which the profiled person is construed.