In this presentation a classic narrative—Defoe’s Robinson Crusoe—will be seen as webular map. That is, rather than reading the story in a linear fashion, it will be read as a dynamic system represented as a multi-branching semantic web. The web is a conceptual representation of the specific semantic families in this particular narrative. The web also represents a kind of ideal memory of the text. That is, if we can imagine how words “end up in the brain” to be filed somehow, so as to be remembered, they might take the shape of a web, spelling out sets of relationships that are not apparent in the naked linear text, or to the ordinary memory.

When the trained (and untrained) wetware of the brain connects to the computer hardware and software at our disposal, a cyborg emerges capable of being a prodigious reader: able to remember every instance of every word and its location; able to create structures to keep organized data from hundreds or thousands of novels; and also able to make nuanced categorical decisions, based on evidence and nuances bred into intuition; and able to think about what all this might mean.

In concrete terms, the presentation will show thirty maps that open Robinson Crusoe’s semantic matrix. The fundamental base of the analysis is the idea that the text is comprised of three zones—the human being/body, the built world, and the raw universe. In this schema: the human in addition to intra-action and interaction, transforms the raw universe and makes and lives in the built world. Being able to see all the constituents of these zones along with their dynamics and qualities, allows us to consider the politics, psychology, ecology, sexuality. and economics (for example) of texts as evidence-rich medium.

The presentation is visual, a demonstration of what we might call graphic criticism. The massive web, surely a reflection of the neurologic, is the dream of new criticism’s desire for close readings, as it is as well of the structuralist idea of seeing the synchronic.

The radical stylistic difference between the early and late novels of Henry James is well known, and, as I have shown elsewhere, the frequencies of the most frequent words do a remarkably good job of characterizing the gradual and unidirectional evolution of his style from the novels of the 1870’s to those of the early 1900’s. This paper, inspired by a classic study by John F. Burrows in 1992 (“Not Unless You Ask Nicely: The Interpretative Nexus Between Analysis and Information”, LLC 7: 91–109), investigates the extent to which James’s other writings mirror the style evolution of the novels. After extracting the most distinctively early and late words in his seven earliest and six latest novels, I analyze all of his remaining novels and a large selection of short fiction on the basis of these same words. Although these texts played no part in the construction of the word lists, the distinctively early and late words show the same gradual and unidirectional development. The same is largely true of James’s non-fiction, letters, and drama, though the differences in genre and in the amount of text available in these categories require some methodological innovations. What emerges is an almost astonishingly regular pattern of cross-genre chronological evolution in the style of Henry James. Whether or not the styles of other authors display the same regularity remains to be seen.

This paper demonstrates computational analysis of both keyword and collocation occurrence in Beerbohm’s two most famous narrative works: Zuleika Dobson—his only novel—and Seven Men—his satiric portrait of seven vain men, Beerbohm himself being one these seven men. For this work I created two comparison corpora, one roughly six times the size of the Beerbohm corpus (for keyword analysis) and one approximately the same size as the Beerbohm corpus (for collocation analysis). The analyses result in a discovery of an abnormally high concentration of past perfect verb complexes in Beerbohm’s works. Further, within those complexes, the verbs known and seen, the latter with both perceptual and cognitive meanings, are more common in Beerbohm’s texts than in the comparison corpora. The distributions of these cognition/perceptual verbs in past perfect complexes not only provide insight into Beerbohm’s style but also allow us to explore the past perfect in one of its primary functions posited by discourse analysts: providing background for foregrounded narrative.