Creating a separate node lets the author zoom in on the painting. Note how this new close-up
view is accompanied by even more specific description. The author makes note of the "flared nostrils and
protruding tongue." Even more promising may be the way the study's sharpened
description prompts the author to refine what was in the general description
the essence of emotions into the fleshed-out scream of the horse. To be fair, the sample has its faults. For one thing, it is too reliant on the previous node which discussed the painting in general. The original sample's movement toward individual figures is implicit in the horses' opening description. However, the nodal nature of hypertexts in most cases calls for more explicit presentation of information and reiteration of relationships and contexts. It's possible that a reader arriving at the horse node may not have previously read the general description. In this case, she would likely be befuddled by the description's reference to multiple figures in a picture completely dominated by the horse's description. The second sample is much like the dependent clause: without the support of the general description, it is but a fragment.
The failure to fully refine the second sample can be addressed by pushing hypertextual nodes toward self-reliance. This should not mean, however, that we should lose track of the interrelations that exist between nodes. Picasso's use of distortion, for example, is mentioned in both samples. Unfortunately, neither descriptions makes an explicit connection with the other on these terms. Perhaps, the kind of work we might ask a hypertextual node's introduction to do might parallel the transition sentence or section. Hopefully an author will make relevant connections and announce the direction the project is about to take.
