*Associative nets* is an empirical term. It refers to something which can be demonstrated or illustrated. The illustration used in the earlier blog is particularly useful in showing what this concept refers to and for this reason is repeated here!
One can select any number of terms in our everyday vocabulary and draw a network of how each term is (or is not) related to some (but not all) other terms of the language. By the same token, if you select any term within an extant language you can show to what other terms in that language it is related. One can use distance, as well as the size of the area covered by a term, or the direction of the connecting link to “map” the relationship between different term. However the nets are multi-directional; the space is n-directional. I add another feature: temporality, that is that nets develop over time so that a shift in the net may influence large areas, a kind of ripple effect which spread from an impact point to distant places, yet does not extend over all surfaces.
If this image of an associative net is moderately correct, it should be clear that it is almost impossible to treat a language-in-use as a stable unit. Such a language deserves to be viewed as a river into which one cannot step twice, yet for practical purposes one views it as the same river. If one builds a dam in the river to store its waters, then later converts the dam to also serve as a source of power, the river retains its name and even most of its contours from source to end. One may revise one’s private image of the river, but this is optional and depends on the situation itself. In fact, we are aware that matters have changed and we accommodate our thoughts accordingly. We use our language not as an object made in heaven, but as a flexible tool which will serve our purposes.