248
R. FISCHER
The Sensory Steady States of Learning and Perception
Evolution, we have shown, is a steady state represented by a straight
line on a log-log plot relating metabolic body size (space) and lifespan
(time). By analogy, learning, another adaptational process, is also a steady
state, using less energy, proceeding at a faster rate of change. Learning can
also be characterized by higher complexity and less stability in energetic
terms or, in other words, it is a less space-like and more time-like step than
evolution. Notwithstanding these differences, the overriding similarity of
the two should result in cumulative learning curves which also follow a
straight line in log-log coordinates since they can be described by power
functions of the general form:
(3)
where P is a cumulated measure of performance and t is practice time or
number of trials [37]. The similarity between this and equation (1), HUXLEY'S
[19] allometric relation pertaining to evolution, is quite apparent. From
the many cumulative learning curves in log-log coordinates [37] we have
selected one example as an illustration in Figure 2, labelled "learning";
tl
2'
c
:J
C I
C
::),
.~
CI
~I
.0
-=:1
~
.01 ~
L,
u
(fJ 1
(fJ I C
C
C I ' "
01
0 I D' "
D- I (fJ
E
U) '
""iii U
I -'
U
I
en
0
I
(5
:J L
.L
E
u
>,
:J
U)
U
D...
1000
100
10
/
/
(:>,/
~<::/
r?;/
-$'/
/
/
/
/
1·
001
01
10
Metabolic body size (kg3/4)
1------,0------10 L 6-----lCio6--LAlTlount .. 9fpractl~~,.(ar~ltrary,unlts) ..
1
10
100
1000
Stimulus magnitude (arbitrary units)
100
1000
Fig. 2. Display of similarity between evolution, learning and perception as
steady states. The straight line plot labelled "evolution" is the interspecies regression line from Fig. 1 and is replotted here only for comparison. The line labelled
"learning" is a replot of KIENTZLE'S data [23], whereas, the line labelled "perception" is replotted after [39]
it illustrates a writing-upside-down task with the abscissa containing the
amount of practice from 1-70 min and the ordinate, the number ofletters
[23]. For comparison, Figure 2 also recapitulates the "interspecies generalization" of Figure 1 labelled "evolution".
R. FISCHER
The Sensory Steady States of Learning and Perception
Evolution, we have shown, is a steady state represented by a straight
line on a log-log plot relating metabolic body size (space) and lifespan
(time). By analogy, learning, another adaptational process, is also a steady
state, using less energy, proceeding at a faster rate of change. Learning can
also be characterized by higher complexity and less stability in energetic
terms or, in other words, it is a less space-like and more time-like step than
evolution. Notwithstanding these differences, the overriding similarity of
the two should result in cumulative learning curves which also follow a
straight line in log-log coordinates since they can be described by power
functions of the general form:
(3)
where P is a cumulated measure of performance and t is practice time or
number of trials [37]. The similarity between this and equation (1), HUXLEY'S
[19] allometric relation pertaining to evolution, is quite apparent. From
the many cumulative learning curves in log-log coordinates [37] we have
selected one example as an illustration in Figure 2, labelled "learning";
tl
2'
c
:J
C I
C
::),
.~
CI
~I
.0
-=:1
~
.01 ~
L,
u
(fJ 1
(fJ I C
C
C I ' "
01
0 I D' "
D- I (fJ
E
U) '
""iii U
I -'
U
I
en
0
I
(5
:J L
.L
E
u
>,
:J
U)
U
D...
1000
100
10
/
/
(:>,/
~<::/
r?;/
-$'/
/
/
/
/
1·
001
01
10
Metabolic body size (kg3/4)
1------,0------10 L 6-----lCio6--LAlTlount .. 9fpractl~~,.(ar~ltrary,unlts) ..
1
10
100
1000
Stimulus magnitude (arbitrary units)
100
1000
Fig. 2. Display of similarity between evolution, learning and perception as
steady states. The straight line plot labelled "evolution" is the interspecies regression line from Fig. 1 and is replotted here only for comparison. The line labelled
"learning" is a replot of KIENTZLE'S data [23], whereas, the line labelled "perception" is replotted after [39]
it illustrates a writing-upside-down task with the abscissa containing the
amount of practice from 1-70 min and the ordinate, the number ofletters
[23]. For comparison, Figure 2 also recapitulates the "interspecies generalization" of Figure 1 labelled "evolution".