Havard R. Glattre, Codestitcher AS, Oslo, Norway
Eystein Glattre, Norwegian School of Veterinary Science, NMBU, Norway

Abstract: This paper is about how fractal connections between words can be identified and hidden fractal word-networks uncovered in corpus texts. The basis for this search-and-find procedure is the hypothesis Complexity Connections in Corpuses (CCIC) consisting of four separate sub-hypotheses, each the basis of a procedure for uncovering a special part of such networks. The validity and applicability of CCIC are demonstrated by showing that each sub-hypothesis holds good for twenty-four masterpieces, selected from the world s great library of novels and treatises. And the hidden structure is developed by means of our computer-program ZIPPF. It turns out that this hidden structure is similar for all investigated corpus texts, a fact which, in our thinking, supports the idea that the hidden word-network renders us a glimpse of working processes used by the brain machinery to produce text. The conclusion is that CCIC provides tools for linguistic research, is a supplement to linguistic networks theory, and opens up for, inter alia fractal-mathematical, studies of high-level brain/mind activity because it follows from CCIC that human verbal thinking is fractal.

Keywords: CCIC-hypothesis, inter-word distance, pink word, white word, a-distribution, self-organization, fractal network