Generating Additional Matrices of Continua

Generating Additional Matrices of Continua
This analysis suggests that the cognitive matrix of continua depicted in Fig. 1
is a specific example of a more general matrix from which examples such as
that of Fig. 1 can be generated. If so, that more general matrix should be
capable of generating not only the psychological example of Fig. 1, but also
a specific example applicable to evolutionary biology. The ability to
generate a general matrix from Fig. 1 and to generate, in turn, an example
applicable to evolution would provide evidence for the argument that
common information structures underlie human cognitive architecture and
evolution by natural selection. Figure 2 depicts a general matrix of continua
that can be used to generate specific matrices applicable to particular areas
that may have the same underlying information structures. Figure 3 depicts
the evolutionary example that can be derived from Fig. 2.
The first continuum of Fig. 1 deals with learning. On the left side of this
continuum, we need to learn (or adapt) when we do not have knowledge
240 John Swellerneeded to function in a particular environment. On the right side, essential
knowledge has been acquired. In the more general terms of Fig. 2, on the left
side, the first continuum deals with an information system that is operating
in a novel context for which it is poorly adapted. It needs to adapt or
Fig. 2. A generalized matrix of continua.
Evolution of human cognitive architecture 241‘‘learn.’’ On the right side, the system has already adapted or ‘‘learned’’
what is needed to operate in its environment. The first continuum of the
specific evolution by the natural selection continuum of Fig. 3 varies from
Fig. 3. A matrix of continua for evolution by natural selection.
242 John Swellerorganisms that are poorly adapted to their current environment and so need
to adapt to organisms that are well adapted to their environment.
The second continuum of each of the three figures is concerned with the
extent to which performance is guided by established rules. In the case of
Fig. 1, dealing with cognitive architecture, on the left side when faced with
new material, there are no schemas to guide performance. On the right side,
when dealing with familiar material, schemas determine actions. Thus, in the
general terms of Fig. 2, on the left there are no available rules to govern the
way the system should operate in its environment, whereas on the right there
are well-established rules. This general continuum is the second continuum
of Fig. 2. Translated into evolutionary terms, on the left we have a genetic
endowment that will not permit a species to survive without change, whereas
on the right we have a species with a genetic endowment that is well adapted
to the current environment.
If a system is not adequately adapted to its environment, it needs to alter.
The left side of the third continuum of Fig. 1 indicates that humans engage
in problem solving when faced with such a situation. On the right, where
material is well learned, adaptation or problem-solving search is unneces-
sary. The third continuum of Fig. 2 describes a general continuum in which
at one extreme, many new procedures are required to permit the system to
operate in the prevailing environment to a situation at the other extreme
where no new procedures are required because the system is well adapted to
the current circumstances. Similarly, in the genetic terms of the third
continuum of Fig. 3, many alterations to the genome are required for
survival on the left side of the continuum as opposed to no requirement for
alterations to the genome on the right side.
If change is required, what are the mechanisms of change? For human
cognitive architecture, the left side of the fourth continuum indicates that
change occurs randomly. (Recall that while the generation of possible
changes is random, assessment of the eVectiveness of possible changes is not
random.) On the right side of the continuum, change is not required because
previously acquired schemas indicate what actions to take faced with a
problem. In other words, we have a system that must generate new
procedures randomly and test them for eVectiveness at one extreme of the
fourth continuum of Fig. 2 or is able to use currently established procedures
at the other end of the continuum. In evolutionary terms, as depicted in the
fourth continuum of Fig. 3, random mutation and sexual recombination are
needed to generate changes to the genome and perhaps new species if a line
is to survive. Alternatively, at the other end of the continuum, the current
genome is satisfactory for survival without substantial alteration.
Finally, if elements are combined randomly, there must be mechanisms
that ensure combinatorial explosions are kept in check. The limited working
Evolution of human cognitive architecture 243memory on the left side of the fifth continuum of Fig. 1 provides such a
mechanism. In contrast, on the right side of the continuum, working
memory limitations are not needed and do not occur because previous
learning has ensured orderly and appropriate sets of elements irrespective of
the size of those sets. In general terms of the fifth continuum of Fig. 2, if new
procedures are being generated randomly, there must be mechanisms to
limit their complexity. Changes must be relatively small and simple to
reduce the number of possible changes and to reduce the probability that
any change will result in a breakdown of the system. On the right side of the
fifth continuum, procedures that are eVective need have no limits to their
complexity. In other words, while changes to the system must be small and
incremental, there are no limits to the complexity of the resulting system.
From the perspective of evolution by natural selection, while alterations to
the genome from one generation to the next are minimal, as indicated on the
left side of the fifth continuum of Fig. 3, that process, if permitted to
continue for a suYciently long period, can result in the immensely complex
genome referred to on the right of the fifth continuum. There may be no
limit to genetic complexity under such circumstances.
The isomorphism of Figs. 1, 2, and 3 provides evidence for the suggestion
that human information processing recapitulates evolution by natural
selection. They both share common information structures. It is understand-
able that the management of information by human cognitive architecture
and evolution by natural selection should be similar. Evolution by natural
selection is possibly the most eYcient, natural system for transmitting,
altering where necessary, and perpetuating information. It might be
expected that human cognitive architecture, which must also manage
information, would evolve to mimic the information processing procedures
of evolution by natural selection because both systems are based on the
general information processing procedures of Fig. 2