Comparative analysis:
Basic concepts
Comparative analysis of learning and cognition
Sponges
Cnidarians
Flatworms
Arthropods
Vertebrates
Mollusks
Nematods
Annelids
•Neurons
•Radial symmetry
•CNS
•Bilateral symmetry
•Ganglionar
nervous
system
•Cranium/Brain
Taxonomic context
Key issues in the comparative analysis of learning and cognition
Issue # 1: distinguish between learning phenomena and learning
mechanisms
Issue # 2: homology of learning mechanisms is demonstrated when
equivalent phenomena are shown to depend on the same
processes at all levels of analysis.
Issue # 3: divergence in learning mechanisms is demonstrated when
species differences in learning phenomena cannot be attributed to
contextual variables.
Key issues in the comparative analysis of learning and cognition
Issue # 1: distinguish between learning phenomena and learning
mechanisms
Issue # 2: homology of learning mechanisms is demonstrated when
equivalent phenomena are shown to depend on the same
processes at all levels of analysis.
Issue # 3: divergence in learning mechanisms is demonstrated when
species differences in learning phenomena cannot be attributed to
contextual variables.
Psychological
Neurobiological
Neurochemical
Cell-molecular
Mechanistic Level Example Modular Representation
S→S learningAPPETITIVE FEAR
Coincidence
detectorsCEREBELLUM
S S S S
AMYGDALA
Synaptic
receptors
SPECIFICITY
NMDA GABA
Second-
messenger
systems ARACHIDONIC ACIDcAMP
Issue # 1: distinguish between learning phenomena and learning mechanisms
Key issues in the comparative analysis of learning and cognition
Issue # 1: distinguish between learning phenomena and learning
mechanisms
Issue # 2: homology of learning mechanisms is demonstrated when
equivalent phenomena are shown to depend on the same
processes at all levels of analysis.
Issue # 3: divergence in learning mechanisms is demonstrated when
species differences in learning phenomena cannot be attributed to
contextual variables.
Coover et al., 1973, J Comp Physiol Psychol, 85, 111-122.
Issue # 2: homology of learning mechanisms is demonstrated when equivalent
phenomena are shown to depend on the same processes at all levels of analysis.
•Lesions in the basolateral
nucleus of the amygdala.
•Sham: simulated lesion.
•Two-way active avoidance
training.
Avoidance learning in rats
Issue # 2: homology of learning mechanisms is demonstrated when equivalent
phenomena are shown to depend on the same processes at all levels of analysis.
Avoidance learning in goldfish
Master-yoked design Impaired by MP but no LP lesions
MP: medial pallium (amygdala homologous)
LP: lateral pallium (hippocampus homologous)
Homology or homoplasy?
Some learning phenomena found in mammals and honeybees
Overlearning extinction effect
Partial delay of reinforcement extinction effect
Overshadowing
Potentiation
Summation
Within-compound associations
Compound uniqueness
Conditional discrimination
Successive negative contrast
Partial reinforcement extinction effect
Positive behavioral contrast
Progressive improvement in spatial reversal learning
Dimentional transfer
Second-order conditioning
Spontaneous recovery in extinction
Latent inhibition
US preexposure effect
Escape and avoidance learning
Bitterman, 1988, Vertebrate-invertebrate comparisons.
Rats Honeybees
>600 Mya
(a planaria-like animal)
Key issues in the comparative analysis of learning and cognition
Issue # 1: distinguish between learning phenomena and learning
mechanisms
Issue # 2: homology of learning mechanisms is demonstrated when
equivalent phenomena are shown to depend on the same
processes at all levels of analysis.
Issue # 3: divergence in learning mechanisms is demonstrated when
species differences in learning phenomena cannot be attributed to
contextual variables.
Determinants of performance
Acquired
behavior
Learning
factors
Contextual
variables
Signal/Response
strength
Memory retrieval
Sensory/Perception
detection
Motivation
Motor control
Issue # 3: divergence in learning mechanisms is demonstrated when species
differences in learning phenomena cannot be attributed to contextual variables.
Two views of divergence
Ecological view:
•Learning mechanisms solve specific ecological problems
•Species vary in their ecology
•Thus, divergence in mechanisms should be widespread
•And associative selectivity should be common
Examples:
•Taste aversion learning seems to violate equipotentiality
•Imprinting and song learning occur during a sensitive period
•Some species show selectivity for a specific sensory dimension
Two views of divergence
General-process view:
•Most ecological niches involve causality, space, and time
•Learning mechanisms deal with such common dimensions
•Thus, divergence in mechanisms should be rare
•And similar phenomena should appear in very different lineages
Examples:
•Contextual factors account for many examples of divergence
•Conditioning occurs in all animals with a CNS
•Conditioning may be involved in some otherwise specialized
cases
Morphological divergence in Hawaiian honeycreepers
Do learning
mechanisms diverge
like the beaks of
these honeycreepers?
Or are they as
conservative as the eyes
or the feathers of these
honeycreepers?
Case study:
Surprising nonreward
Comparative analysis of learning and cognition
Osteichthyes
(Goldfish)
Amphibia
(Toads)
Aves
(Pigeons)
Reptilia
(Turtles)
Mammalia
(Rats)
Phylogenetic relationships among fish, amphibians, and reptiles
Classic learning theory
Thorndike, 1898, Psychol Rev
Strengthening-weakening mechanism
Stimulus Response … Incentive
Stimulus Response … Nothing
Thorndike, 1911, Animal ingelligence
Emotional (and cognitive) factors
Tinklepaugh, 1928, Calif Pub Psychol
Drawing by Katsuo & Chiharu Tomita
Elliott, 1928, Univ Cal Pub Psychol
Successive negative contrast (SNC)
SNC
Sunflower seeds
Bran mash
Sunflower seeds
(for both groups)
Preshift Postshift
Group 1-9 10-16
Downshifted (9) Bran mash (6) Sunflower seeds
Unshifted (9) Sunflower seeds (6) Sunflower seeds
Consummatory successive negative contrast (cSNC)
Preshift Postshift
Group 1-10 11-15
32-4 32% 4%
4-4 4% 4%
Papini, 2006, Jap J Anim Psychol
(a) Toads
1.0
1.4
1.8
2.2
0 5 10 15 20 25
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(b) Turtles
1.0
1.4
1.8
2.2
0 5 10 15 20 25 30 35 40
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(c) Pigeons
0.0
1.0
2.0
3.0
0 5 10 15
4-Trial Blocks
ln I
niti
al L
ate
ncy
S
L
L-S
(a) Toads
1.0
1.4
1.8
2.2
0 5 10 15 20 25
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(b) Turtles
1.0
1.4
1.8
2.2
0 5 10 15 20 25 30 35 40
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(c) Pigeons
0.0
1.0
2.0
3.0
0 5 10 15
4-Trial Blocks
ln I
nitia
l L
ate
ncy
S
L
L-S
Amphibians (Toads)
SNC
Reversed SNC
(a) Toads
1.0
1.4
1.8
2.2
0 5 10 15 20 25
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(b) Turtles
1.0
1.4
1.8
2.2
0 5 10 15 20 25 30 35 40
Trials
Me
an
Lo
g L
ate
ncy
S
L
L-S
(c) Pigeons
0.0
1.0
2.0
3.0
0 5 10 15
4-Trial Blocks
ln I
nitia
l L
ate
ncy
S
L
L-S
0
2
4
6
8
10
0 4 8 12 16
Trials
Num
ber
of
Err
ors
S
L-S
Mammals (Rats) Birds (Pigeons) Reptiles (Turtles)
0.0
0.5
1.0
1.5
2.0
2.5
0 5 10 15 20 25
3-Trial BlocksM
ea
n L
og
La
ten
cy
S
L
L-S
Bony Fish (Goldfish)
Surprising nonreward
• Cues predict a larger or more preferred reward than that actually occurring.
• Surprising nonrewardpromotes two kinds of learning:
•Allocentric learning: about a change in the environment (cognitive).
•Egocentric learning: about the organism’s reaction to that change (emotional).
Papini, 2003, Brain Behav Evol
Osteichthyes
(Goldfish)
Amphibia
(Toads)
Aves
(Pigeons)
Reptilia
(Turtles)
Mammalia
(Rats)
Phylogenetic relationships among fish, amphibians, and reptiles
?
???? ?
Osteichthyes
(Goldfish)
Amphibia
(Toads)
Aves
(Pigeons)
Reptilia
(Turtles)
Mammalia
(Rats)
Phylogenetic relationships among fish, amphibians, and reptiles
Allocentric learning[Reversed SNC]
Allocentric
+
Egocentric learning[SNC]