HOUSE_OVERSIGHT_013693.jpg

coherent activity. This temporal and phase coherence plays an important role in
current theory of sensory-associative-motor integrative function, how distributed
attributions come together in the brain representation of a single object or process,
in the context of the so-called “binding problem” (Singer, 1993; Bressler, 1995;
Nicolelis, 1995; Schiff et al, 1997). Diffusely distributed neurochemical variables
have been invoked. For example, the role of metabotropic glutamate receptors in
driving the synchronization of interneuronal networks has been suggested as a
mechanistic model (Whittington et al, 1995). The objects of relevance to the
discovery of Fermi-Pasta-Ulam are studied as the nonlinear physics of
nondissapative wave processes and are called solitons (Zabusky and Kruskal,
1965). They have been invoked to model nerve conduction and information
transport in brain (Scott, 1990).

A third counter-intuitive set of accidental computational findings is in an area
of research called symbolic dynamics which involves the universal parameter-
dependent coding language and capacity of nonlinear systems. In the early 1960’s,
a group around Stan Ulam at Los Alamos (Cooper, 1987) used one of the early
“high powered” computers, MANIAC II, to iterate (letting the output of the action of a
discrete time function serve as its input the next time around) simple equations they
called “maps.” These reduced dimensional objects shaped like tents, sine functions
and parabolas can be extracted from and represent the behavior of higher
dimensional, nonlinear differential equations (see Devaney, 1989; Schuster, 1989 or
Moon, 1992 for intuitive descriptions). They varied a parameter, such as the height
of the tent or parabola, to systematically change the period and/or phase (order) of
the symbol sequence (Metropolis et al, 1973). Normalizing the range of values of
the output to [0,1] and transforming the series of values into a binary code, L ≤ 0.5
and R > 0.5, they found an invariant, one parameter dependent, progression of
ordered periods, R, RLR, RLRR…RLLRL…, in all such single maximum maps. This
“U (universal)sequence” has also been found as singly or multiply present in a
variety of real systems, including complicated chemical reactions (Simoyi et al,
1982; Coffman et al, 1986). This means one can “dial” the parameter to generate
“words” of sufficient computational complexity to serve as a language. These

193

HOUSE_OVERSIGHT_013693