How Information Structures Have Impelled the Evolution of Human

How Information Structures Have Impelled the Evolution of Human
Cognitive Architecture
A. Information Structures
While considerable work by many researchers over several decades has been
devoted to the organization of human cognitive architecture, far less eVort
has gone into investigating the information structures that must have driven
the evolution of that architecture. Some work has been carried out by
Sweller (1994) and Halford, Wilson, and Phillips (1998). Sweller (1994)
suggested that all information can be placed on a continuum according to
the extent to which the elements that constitute the information interact. At
one extreme, there is no interaction between the elements that need to be
learned. They are independent. Element interactivity is low or, indeed,
nonexistent, which means that each element can be considered and learned
serially without reference to any other element. Because elements at the low
element interactivity end of the continuum do not interact with each other,
there is no loss of understanding despite each element being learned
individually and in isolation. Understanding is defined as the ability to
process all elements that necessarily interact simultaneously in working
memory. Learning such material imposes a low cognitive load because each
element can be learned without reference to other elements.
At the other extreme of the continuum, there is close interaction between
the various elements that need to be learned. Element interactivity is high,
which means that if the material is to be understood, all of the information
with its multiple elements must be processed simultaneously, imposing a
heavy cognitive load. Elements that interact can be processed individually,
in serial fashion, but not with a high degree of understanding. Processing
high element interactivity material without learning necessary relations
between elements will result in rote learning. The reason rote learning occurs
216 John Swellerfrequently is because learning individual elements without learning
important relations and interactions between elements can reduce cognitive
load dramatically. When rote learning, only one, or at most, a very limited
number of elements need to be held or processed simultaneously. In eVect,
during rote learning, high element interactivity material is treated by the
cognitive system as though it is low element interactivity material.
In contrast, learning high element interactivity material with understand-
ing imposes very heavy cognitive demands, especially if there are many
interacting elements. For understanding to occur, all interacting elements
must be processed simultaneously, and for some extensive, high element
interactivity material, processing all of the interacting element simultan-
eously may be very diYcult or even impossible (Pollock, Chandler, &
Sweller, 2002). Learning such material by rote reduces cognitive load, but at
the cost of understanding. Examples of very low and very high element
interactivity material are discussed next