Transcript Chapter 1

大學部 生態學與保育生物學學程 (必選)
2010 年 秋冬
學習與認知
(Learning and Cognition)
─動物行為學 (Ethology)
鄭先祐(Ayo)
國立 臺南大學 環境與生態學院
生態科學與技術學系 教授
Ayo NUTN Web: http://myweb.nutn.edu.tw/~hycheng/
Part 1. 動物行為的研究途徑 (個體行為)
 歷史背景 (History of the Study of Animal




Behavior ).
基因分析 (Genetic Analysis of Behavior ).
天擇 (Natural Selection and Behavior ).
學習與認知 (Learning and Cognition.)
生理分析 (Physiological Analysis)


(一) 神經細胞 (Nerve Cells and Behavior ).
(二) 內分泌系統 (The Endocrine System).
 發育(The Development of Behavior ).
Ayo 教材 (動物行為學 2010)
2
05 學習與認知 (Learning and cognition)
 Definition of Learning
 Types of Learning
 Habituation (習慣)
 Classical conditioning
 Operant conditioning
 Latent Learning
 Social Learning
 Species differences in Learning
 Comparative Studies
 Other Evidence of Cognitive abilities in animals
Ayo 教材 (動物行為學 2010)
3
People have presuppositions about the mental
lives of animals
 We discount the abilities of some species
 Especially those very different than us
 We anthropomorphize our pets and other primates
 We assume that they think like we do
 How do we know what animals know?
 Does a chimpanzee plans its actions to catch termites?
 Does an ant lion(蟻獅) understand its own behavior
Ayo 教材 (動物行為學 2010)
4
Definition of Learning
 It is “a process through which experience changes an
individual’s behavior”
 Learning is a change in our capacity for behavior as a
result of experience

Excluding the effects of fatigue, sensory adaptation, or
maturation of the nervous system
 Behavioral changes resulting from learning are not
always expressed immediately

We can’t know whether an animal has learned
something just by seeing a change in its behavior
Ayo 教材 (動物行為學 2010)
5
A type of learning: habituation
 In habituation, the animal learns not to respond to a
particular stimulus

Because the stimulus has proven to be harmless
 Habituation: the waning of a response after repeated
presentation of a stimulus.

Once it occurs, its effects are long lasting
 Habituation is everywhere, from protozoans to
humans

It is often considered the simplest form of learning
Ayo 教材 (動物行為學 2010)
6
Habituation in the clamworm (沙蠶)
 雙齒圍沙蠶（Nereis
succinea）是一種廣泛分佈
的多毛綱動物。
 It partially emerges from its
tube to filter tiny bits of food
from the water



withdraws into its tube when it senses danger (i.e. a shadow)
In the lab, subsequent presentations decreased escape
responses
The clamworms had habituated

The effects lasted several hours
Ayo 教材 (動物行為學 2010)
7
 Habituation of
the withdrawal
response to a
shadow by the
clamworm.
Ayo 教材 (動物行為學 2010)
8
Habituation is not due to adaptation
 The clamworms’ decline in responsiveness was not
because the sense organs became adapted to the
stimulus

Nor was the decline due to muscle fatigue, because
habituated worms still withdrew in response to
prodding
 The clamworms had learned to stop responding to the
shadow
Ayo 教材 (動物行為學 2010)
9
Habituation is specific to a particular
stimulus
 Young turkeys, chickens, and pheasants innately
show antipredator behaviors


They crouch and give alarm calls at the sight of
objects moving overhead
When a model was presented frequently, it elicited
fewer and fewer alarm calls from chicks
 As adults, they respond only to the image of a
predator

Such as a hawk flying overhead
Ayo 教材 (動物行為學 2010)
10
Habituation also occurs in species
interactions
 Territorial species reduce aggressive responses toward
neighbors, But still respond aggressively toward unfamiliar
intruders
 Bird species and bullfrogs respond aggressively to a recording
of a stranger’s call but not to a familiar call
 Habituation mediates this process
 Animals stop responding to a call when its heard
repeatedly
 The decline was specific to the calls
Ayo 教材 (動物行為學 2010)
11
Habituation as an experimental tool
 Habituation is a useful tool to study cognitive processes
in animals, including humans
 A subject is habituated to a stimulus

And then a new stimulus is presented
 If the subject’s response changes
 The experimenter knows it can detect the difference
between the two stimuli
Ayo 教材 (動物行為學 2010)
12
A type of learning: classical conditioning
 Associative learning: an animal makes an association
between a stimulus and a response
 One type of associative learning: classical conditioning
(Pavlov dog)



A dog salivates when powdered food is blown into its
mouth, but not when it hears a bell
These two stimuli are paired immediately before food
powder was presented
The dog salivates in response to the bell alone
Ayo 教材 (動物行為學 2010)
13
Classical conditioning explained
 Unconditioned stimulus (US): an animal has an inborn
response to a certain stimulus

The animal did not have to learn the response
 Unconditioned response (UR): the response to the US
 Conditioned stimulus (CS): a new stimulus is paired
with the US until eventually it, too, can elicit the
response
 Conditioned response (CR): the response to the
conditioned stimulus
Ayo 教材 (動物行為學 2010)
14
Classical conditioning is consistent
 The order of the presentation of the US and CS is
important

The two stimuli must occur close together
 Useful signals are reliable
 They predict that a particular event or stimulus will
follow
 An association between a CS and US can be lost
 If the CS is no longer reliable, the subject stops
responding
 Extinction: the loss of the conditioned response
Ayo 教材 (動物行為學 2010)
15
The adaptive value of classical
conditioning
 Learning through classical conditioning provides
fitness advantages to wild animals
 Few studies have addressed the value of classical
conditioning in the everyday life of an animal



The territorial and reproductive behaviors in blue
gouramis (fish that inhabit shallow pools and streams in
Africa and Southeast Asia)
Sperm production in male field crickets
Feeding behavior in honeybees
Ayo 教材 (動物行為學 2010)
16
案例：the blue gourami (三星鬥魚)
 A male blue gourami defends its territory aggressively
 The intruder may respond with a submissive posture or
retreat
 If not, the contest escalates into a battle that can result in
serious injury
 Dangerous fights evolve when the value of the resource is
great
 Success is crucial for males because females won’t mate
with a male without a territory
Ayo 教材 (動物行為學 2010)
17
Blue gourami (三星鬥魚) learning
 With victory, a male blue gourami gains an
immediate competitive edge and experience

Increasing the probability of winning future battles
 If a male could learn the signals (visual, chemical, or
mechanical) that indicate the approach of a rival

He might be better prepared for battle and gain a
competitive edge
 Males that had been classically conditioned to
associate a light with the imminent appearance of a
rival were superior in territorial defense
Ayo 教材 (動物行為學 2010)
18
The “winner effect”
 Successful males are more likely to attract females
 But excessive aggressiveness harms mating success
 A male conditioned to expect the arrival of a female is
less likely to attack her (研究者使用「光」)
 Classical conditioning pays off in reproductive success
 In nature, other cues (i.e. the shape of a gravid belly) for
may be a reliable cue of a willing female’s approach.
Ayo 教材 (動物行為學 2010)
19
Conditioned
blue gourami
males have
higher
reproductive
success
Ayo 教材 (動物行為學 2010)
20
案例：male field crickets
 Female field crickets mate more than once
 Sperm from different males compete inside her body to
fertilize her eggs
 A male increases his chances of fathering more
offspring by transferring more sperm to the female
 Males can learn to associate environmental cues with
the presence of male competitors
Ayo 教材 (動物行為學 2010)
21
 Field crickets learned
about spatial cues that
signaled the presence of
a competitor.
 Male produced larger
sperm packets in the
environment associated
with a competitor.
Lego bricks
Ayo 教材 (動物行為學 2010)
22
案例：feeding honeybees
 A foraging Bee can be rapidly conditioned to respond
to odor



When the antennae of a bee are touched with a sucrose
solution (the US), the bee extends its proboscis to lick it
(the UR)
When an odor is presented just before the sucrose
solution is presented, the bee rapidly forms an
association between the odor (the CS) and the sucrose
And begins to extend its proboscis to the odor alone
Ayo 教材 (動物行為學 2010)
23
Operant (instrumental) conditioning
 Another form of associative learning
 It is also called trial-and-error learning
 When cats inside a “puzzle box” accidentally hit a lever in
the correct way, the door would open, and the cat would get
to eat

Over successive trials, a cat would get faster and faster at
performing the correct behavior to release the latch
 This type of learning emphasizes that the animal operates
on the environment to produce consequences
Ayo 教材 (動物行為學 2010)
24
The Skinner box
 Easier to use than
Thorndike’s puzzle box

It’s still used today
 A hungry animal inside the
Skinner box must learn to
manipulate a mechanism
(pressing a lever or
pecking a key)

To get a food reward
Ayo 教材 (動物行為學 2010)
25
Reinforcers
 Reinforcer: a stimulus, such as a bit of food, that changes the
probability that an animal will repeat its behavior
 Positive reinforcer: increases the probability of a behavior being
repeated, such as food offered to a hungry rat or a drink to a
thirsty one
 Negative reinforcer: increases the probability of a response once
it is removed
 If an unpleasant or painful stimulus stops when an animal
performs a certain act, it is likely to repeat that action
 It’s different than punishment: a decrease in a response due to
the presentation of an aversive stimulus
Ayo 教材 (動物行為學 2010)
26
Timing of events is critical
 In operant conditioning and classical conditioning
 When the animal spontaneously performs a behavior,
reinforcement must follow closely
 When reinforcement is withheld, the response rate
gradually declines and become extinguished

Just as the strength of the conditioned reflex decreases
when the CS is presented many times without the US
Ayo 教材 (動物行為學 2010)
27
Shaping (導引化)
 Operant conditioning can be used to teach animals to
perform novel and complex acts
 Shaping: used by Hollywood animal trainers


At first, the trainer reinforces any approximation of the
desired act
Later, it requires better performances to get a reward
 To train a sea lion to jump through a hoop, first reward it
for approaching the hoop



Then reward it only when it swims through the hoop
Raise the hoop on successive trials
Offer the sea lion a fish only when it makes the leap
Ayo 教材 (動物行為學 2010)
28
Reinforcement(增強作用) schedules
 Realistically, the reward does not follow every
performance of an act
 Reinforcement schedule: the frequency with which
rewards are offered
 Reinforcement schedules can change
 Each reinforcement schedule has predictable effects


On the rate of response and
On how long the animal continues responding when it is
no longer rewarded
Ayo 教材 (動物行為學 2010)
29
Reinforcement schedules
 A continuous reinforcement schedule: each occurrence
of the behavior is rewarded
 Is best during the initial training to establish a response
 A fixed ratio schedule: the animal must respond a set
number of times before being rewarded
 Very high response rates because the individual
determines how quickly it will be rewarded
 A variable ratio schedule: the number of responses
required for reinforcement varies randomly
 The individual is rewarded for fast responses
 The response persists even if the reward is withheld
Ayo 教材 (動物行為學 2010)
30
Latent(潛伏) learning
 Sometimes animals learn without any obvious reward
 When animals learn important characteristics of their environment
during unrewarded explorations, they can use this information later
 Familiarity with the terrain(地形) improves survival
 Knowledge of the environment helps them evade predators
 Even ants gather information for later use
 Ants evaluate prospective nest sites based on a range of criteria:
floor area, headroom, entrance size, darkness, hygiene, and the
proximity of hostile neighbors
Ayo 教材 (動物行為學 2010)
31
Ants keep track of local housing options
 If a nest is destroyed, ants must quickly find another one
 Ants ignore a familiar but unattractive option in favor of
exploring for a better one
 Even if they ultimately settle for something that is also
unattractive
 In latent learning, animals didn’t learn an appropriate
behavior, but they learned something about their
environment

So they could respond appropriately in a new situation
Ayo 教材 (動物行為學 2010)
32
 The experimental
design for a study
of latent learning in
ants.
 Alternative nest #1
was added for a
week. Then a
second nest (#2)
was added.
 Ants choose #2.
Ayo 教材 (動物行為學 2010)
33
Social learning
 Learning from others is not part of every animal’s behavioral
repertoire
 Social species have a greater opportunity for social learning
than do solitary species
 Social learning encompasses a broad range of phenomena
 Some of which suggest a higher level of cognitive skill (認知
能力) on the part of the animal than do others
 Animals may inadvertently provide information to other animals
 Or, individuals share information through specific signals
Ayo 教材 (動物行為學 2010)
34
Social learning: stimulus enhancement
 Information may not be actively communicated by
one animal to another
 In stimulus enhancement: an animal may be attracted
to a particular object because a conspecific is near it or
interacting with it

For example, rats learn dietary preferences(飲食嗜好)
from other rats by smelling their breath
Ayo 教材 (動物行為學 2010)
35
Social learning: local enhancement
 In local enhancement: an animal may be attracted to
a particular location because a conspecific is there
 In nature, local and stimulus enhancement occur
frequently during foraging


Bumblebees (大黃蜂) land on flowers already
occupied by other bees
Other animals also use conspecifics as cues to good
foraging patches
Ayo 教材 (動物行為學 2010)
36
Observational conditioning
 A type of classical conditioning that occurs in social situations
 Animals learn to avoid dangerous situations by watching conspecifics

Rhesus monkeys learn to fear and avoid snakes by watching other
monkeys show their fear
Minnows (米諾魚) learn to show fear responses to pike (狗魚)
odor when they are paired with minnows that had experience with
pike
 Observational learning does not assume that observers understand
anything about the mental state of the animals they are learning from

Ayo 教材 (動物行為學 2010)
37
Goal-directed emulation (仿效)
 An observer seems to learn from observation what goal
is to be achieved, but does not copy precisely what the
demonstrator does
 Chimpanzees and children both watched an adult human
retrieve(回收) artificial fruit from a clear plastic box


Chimps showed goal-directed emulation: they directed
their attention at the correct part of the box but did not
imitate the action of the demonstrator
Children imitated (模仿) the actions of the human
exactly
Ayo 教材 (動物行為學 2010)
38
Imitation (模仿)
 An observer copies exactly what a demonstrator does
 In the two-action test, the subject is presented with a task
that has two equally easy solutions

If subjects choose the solution they have seen demonstrated,
it is evidence of imitation
 Observer budgerigars(鸚鵡, budgies) watched
demonstrator budgies open a dish

The observer used the same technique it had just witnessed
Ayo 教材 (動物行為學 2010)
39
The adaptive value of social learning
 The adaptive value of social learning is clear
 It saves time and energy that might be wasted as an
individual learns the business of survival by trial and
error
 Each individual may have the capacity to learn
appropriate responses

But it is more efficient and less dangerous to learn
about the world from others
Ayo 教材 (動物行為學 2010)
40
Animals learn from each other
 Rats learn about food by smelling each other’s breath and by
observation
 A rat will try novel food if it observes another rat eating it
 Other species may learn routes to food from conspecifics
 Guppies (孔雀魚) learn a safe route by swimming in groups
 Animals also learn from other species
 Group-foraging doves (野鴿) in Barbados learn from other
doves
 But territorial doves learn from Carib grackles, the species
they feed with in mixed flocks
Ayo 教材 (動物行為學 2010)
41
Traditions (傳統)
 Traditions: spread through a group and are stable over
time


In England around 1921, blue tits learned to break into
milk bottles to steal the cream, which floated to the top
This spread through Great Britain as other birds acquired
the habit
Ayo 教材 (動物行為學 2010)
42
Traditions spread through groups of primates
 Young chimpanzees in nature learn to use sticks and
stems to gather termites after watching their mothers
or other adults
 A young female snow monkey discovered that
washing sweet potatoes (provided by researchers) in
the sea cleaned them and enhanced the flavor by
lightly salting it

The tradition spread to others
Ayo 教材 (動物行為學 2010)
43
Individual vs. social learning
 It is hard to tell between individual learning (learning through
one’s own experience) and social learning
 They may occur simultaneously
 For example, sweet potato washing may occur through stimulus
enhancement
 A monkey picks up a dropped potato that has been washed,
like the taste, and then is primed to learn to wash potatoes
 Differential reinforcement may also maintain the behavior
 Human caretakers give more sweet potatoes to members of the
troop that wash them
Ayo 教材 (動物行為學 2010)
44
Comparative studies of species differences
 For decades, the dominant view in studying learning
was that it occurred in essentially the same way across
mammal species

Many characteristics of learning are similar in many
species
 In recent years, researchers have been intrigued not
just by similarities across species, but also their
differences

Several studies documenting differences across species
correlates with the ecological conditions they face
Ayo 教材 (動物行為學 2010)
45
Learned knowledge is not genetically heritable
 In some species, offspring learn from watching their
parents


A jumping spider (跳蛛) can learn that milkweed bugs
(椿象) are not good to eat and ignores them
This knowledge is not passed onto to the spider’s
offspring

Spiders must learn this for themselves
Ayo 教材 (動物行為學 2010)
46
The ability to learn as a heritable trait
 What is heritable, and subject to natural selection, is the
capacity to learn
 The heritability of the ability to learn has been
demonstrated


Fruit flies that had learned the association between
quinine and a particular flavor avoided that flavor, and
laid their eggs on a neutral flavor
After 15 generations, flies from these selected lines
learned the task faster and remember it longer than were
control flies
Ayo 教材 (動物行為學 2010)
47
Learning has costs
 Learning allows an animal to adjust its behavior to new situations.
This ability may not always be advantageous - it has its costs
 It takes time to learn
 Innate behaviors save the time and trouble of making mistakes
 The ability to learn requires neurons dedicated to the task
 Neurons could be devoted to something else (i.e. large olfactory
centers to detect and interpret scents left by prey)
 Learning has an “operating cost”—it takes energy to collect,
process and store information
Ayo 教材 (動物行為學 2010)
48
Species differ in learning ability
 The environment and evolutionary history of a species
influences how learning increases Darwinian fitness

The relative number of offspring left by an individual
 There are biological constraints on learning
 Members of a particular species may be able to learn
certain things and not others
Ayo 教材 (動物行為學 2010)
49
Species differences in learning ability
 Three related species of birds - Clark’s nutcrackers,
pinyon jays, and scrub jays - cache (store) seeds


The birds recover and eat the seeds during winter and
spring when food is scarce
They remember the exact locations, months after they’ve
hidden them
Ayo 教材 (動物行為學 2010)
50
 ←The Clark‘s Nutcracker (Nucifraga
columbiana), is a large passerine bird, in the
family Corvidae (鴉科).
 ↓The Pinyon Jay (Gymnorhinus
cyanocephalus) is a jay between the North
American Blue Jay and the Eurasian Jay in
size.
 ← The Florida Scrub-Jay
(Aphelocoma
coerulescens) is one of the
species of scrub-jay native
to North America..
Ayo 教材 (動物行為學 2010)
51
Bird ecology influences seed storage
 Clark’s nutcrackers cache thousands of pine seeds in
thousands of locations


They live at high elevations with harsh, long winters
They survive almost entirely on stored seeds
 Pinyon jays cache fewer seeds, closer to the collecting
site

Most of their winter diet consists of cached seeds
 Scrub jays store the fewest seeds
 Seeds comprise less than 60% of the winter diet
Ayo 教材 (動物行為學 2010)
52
Seed caching
 Ecological differences among caching species predict that:
 Species that rely on caching to survive the winter have a
better spatial memory
 Nutcrackers and pinyon jays
 Depend most on finding stored seeds to survive the winter
 Did better than scrub jays in finding stored seeds
Ayo 教材 (動物行為學 2010)
53
 Histograms that show the accuracy with which scrub
jays, pinyon jays, and Clark’s nutcrackers find their
caches.
 The solid bar, 15 holes were available for caching.
 The striped bar, 90 holes were available.
Ayo 教材 (動物行為學 2010)
54
Learning is shaped by natural selection
• Clark’s nutcracker was the champion on a spatial task
• There were no species differences in memory on a
nonspatial task
• Remembering the color of a circle was not related to
the species’ dependence on stored food
Ayo 教材 (動物行為學 2010)
55
 Comparison of learning abilities among food-storing
on spatial and nonspatial tasks.
 (a) location of circle
 (b) the color of circle.
Ayo 教材 (動物行為學 2010)
56
 Comparison of learning abilities among food-storing on
spatial and nonspatial tasks.
 (a) location of circle
 (b) the color of circle.
Ayo 教材 (動物行為學 2010)
57
Food storing in other species
 supports the hypothesis that evolution shapes learning ability
is found among other groups of related species
 Titmice and chickadees, a family of birds distinct from
nutcrackers and jays, store seeds and insects
 They performed better than non-food-storing species in tests
 Mammals show a relationship between ecology and spatial skills
 The Great Basin kangaroo rat does not store food, while
Merriam’s kangaroo rat does
 The Merriam’s kangaroo rat performed better on a spatial test
than the Great Basin rat

Ayo 教材 (動物行為學 2010)
58
 The tits, chickadees, and titmice
constitute Paridae(山雀科), a large
family of small passerine birds
which occur in the northern
hemisphere and Africa. Most were
formerly classified in the genus
Parus.
 The Black-capped Chickadee
(Poecile atricapillus) is a small,
common songbird, a passerine bird
in the tit family (山雀科). It is the
state bird of both Maine and
Massachusetts, and the provincial
Ayo 教材 (動物行為學 2010)
59
 Kangaroo rats, genus
Dipodomys, are small
rodents native to North
America.
 Dipodomys microps
Chisel-toothed Kangaroo
Rat
 Dipodomys merriami
Mearns
Ayo 教材 (動物行為學 2010)
60
Evidence of cognitive abilities in animals
 Some scientists wonder whether animals have mental
experiences i.e. thoughts and feelings
 How can we know whether other animals think or
whether they are self-aware?
 Evidence for cognitive abilities(認知能力) includes:




Tool use (工具使用)
Insight (洞察力)
Detours (繞道)
Understanding abstract concepts(抽象概念), including
self-awareness (自我意識)
Ayo 教材 (動物行為學 2010)
61
Tool use (工具使用)
 Tool use: the use of an object in order to obtain a goal
 Once considered a hallmark trait that separated humans
from other animals, tool use is now known in many
species
 For example, a sea otter (海獺) uses a rock to break a
clam shell
 A vulture (禿鷹) drops a rock on an egg, which cracks
open
Ayo 教材 (動物行為學 2010)
62
 Does an animal have to understand how a tool works
in order to use it?


Tool use seems to demonstrate a high level of cognition
(認知)
Animals using a tool to solve a problem often appear to
be thinking it through
Ayo 教材 (動物行為學 2010)
63
Sultan’s tool use
 Sultan learned to use a stick
as a tool to rake in a banana
from outside his cage
 He learned to join two sticks
to reach a banana
 When the sticks separated,
he immediately rejoined
them
 This is evidence that he
understood that joining two
poles increased his reach
Ayo 教材 (動物行為學 2010)
64
Insight (洞察力)
 Insight: a situation where a flash of understanding
occurs

As seen in the suddenness of Sultan’s solution to
obtaining a banana
 The animal may see new relationships among events
 And consider the problem as a whole
Ayo 教材 (動物行為學 2010)
65
 Animals may form a mental representation of the
problem and then mentally work through solutions
 But, all the details of an animal’s prior experience need
to be known to understand what it knows when it
manipulates objects

Seemingly insightful behavior might be specific
stimulus–response relationships learned through
operant conditioning
 To reach meat suspended on a string, a raven (烏鴉)
had to repeatedly pull up a loop of string, step on the
loop, and then reach down and pull up another loop

Some birds can be taught by operant conditioning
Ayo 教材 (動物行為學 2010)
66
Raven (烏鴉)
 Raven is the common
name given to several
larger-bodied members
of the genus Corvus (烏
鴉).
New Caledonia crows→
Ayo 教材 (動物行為學 2010)
67
Insightful behavior in common ravens (烏鴉)
 But, some ravens solved the problem immediately
 Without going through a learning process
 It is unlikely that this complicated behavior was
learned, was genetically programmed, or occurred by
chance
 The ravens apparently have the ability to find
insightful solutions to new problems, using string
as a tool
Ayo 教材 (動物行為學 2010)
68
Insight alone may not always be enough
 Familiarity with a simpler task may be required to
succeed on a new task
 If a string with food was looped up and through the cage,
then down again

The birds had to pull the string down to raise the meat
 Ravens familiar with the pull-up task could quickly do
the pull-down task

But naive birds could not
Ayo 教材 (動物行為學 2010)
69
Animals may understand causal
relationships
 New Caledonia crows make and use tools out of twigs
and leaves
 In the lab, one crow made a hook from a straight piece
of wire to extract a bucket (提桶) from a tube
 Crows may understand that tools can be used to obtain
out-of-reach objects, even other tools
 Video cameras attached to wild crows revealed that
crows keep good tools for future use
Ayo 教材 (動物行為學 2010)
70
Species differ in their ability to understand
 Capuchin monkeys (卷尾猴) learned to push a stick
into a tube to push a peanut out


When a trap was placed into the tube, the monkey had to
push the stick into the correct end
No monkeys fully grasped the task
 Chimpanzees, in contrast, showed more understanding
of the task
 Human children under three years old behave more
like capuchins
Ayo 教材 (動物行為學 2010)
71
Capuchin monkeys
 卷尾猴屬（學名：
Cebus），是卷尾猴
科的一屬，也叫懸猴，
分佈於北起中美宏都
拉斯，南到南美中部
（巴西中部、秘魯東
部和巴拉圭）的狹長
地帶。
Ayo 教材 (動物行為學 2010)
72
 The tube task but with a trap added. The stick must be
inserted in a particular end of the tube. Here, a capuchin
monkey is about to make an error.
Ayo 教材 (動物行為學 2010)
73
 黑猩猩 (Chimpanzees)
Ayo 教材 (動物行為學 2010)
74
Tool use in humans
 Tools are of great importance in human evolution
 We can gain some insight into our own past by
examining the behavior of our close relatives
 Chimpanzees are accomplished tool users
 They use sticks to forage for termites and rocks as a
hammer and anvil to pound open nuts
 Foraging and hunting were probably the first
contexts in which our ancestors used tools
Ayo 教材 (動物行為學 2010)
75
Detours (繞道)
 Detouring: the ability to identify an alternative route to
a reward when the direct route is blocked

Animals can improve on detour tests with experience
 Dogs are not very good at solving detour problems
 often dig under a fence instead of going around it
 Squirrels are very good at solving detour problems
 They choose certain branches to get from tree to tree
Ayo 教材 (動物行為學 2010)
76
 Quail (鵪鶉) and herring gulls easily solved a detour
problem that required them to walk around a barrier,
but canaries (金絲雀) could not

Canaries can fly around an object
Ayo 教材 (動物行為學 2010)
77
How do animals solve detour tasks?
 Portia jumping spiders (跳蛛) excel (勝過) at detours


Their large anterior eyes are specialized for acute vision
They hunt other spiders by climbing into their webs and
luring(誘惑) them with signals similar to those of prey
 Portia can spot spider webs from some distance away
 To reach its prey it must perform a detour
Ayo 教材 (動物行為學 2010)
78
Portia jumping spiders
 Portia is a genus of jumping spider which feeds on other
spiders (araneophagic).
 They are remarkable for their hunting behaviour which
suggests they are capable of learning and problem
solving, traits normally attributed to much larger animals.
Ayo 教材 (動物行為學 2010)
79
Portia spiders use detours
 Before setting out, Portia can
 Choose correctly between detours
that lead to prey versus those that
do not
 Portia looks at the lure(目標), then
slowly scans along the features of
the potential route

If the route ends, the spider turns
back to look at the lure again, and
begins again
Ayo 教材 (動物行為學 2010)
80
Understanding abstract concepts
 Many animals can discriminate relative numbers
i.e. is there more fruit on this tree or that tree?
 The ability to count is more difficult
 To assign a tag such as “1, 2, 3,” to individual quantities
 The ability to count things demonstrates some understanding
of the abstract concept of numbers

Ayo 教材 (動物行為學 2010)
81
Alex, a remarkable parrot
 Alex, an African grey parrot, learned labels (names) for
over 35 different objects

He could identify, request, refuse, or comment on more
than 100 different objects
Ayo 教材 (動物行為學 2010)
82
Alex understood abstract concepts
 He used language to show he understood abstract concepts
 He could:
Say how many items were in a group of up to 6 items,
even if the objects were scattered around a tray
 Count items in a confounded number set (items that vary
in more than one characteristic)
 Add up the total from two sequentially presented
collections
 Alex had some understanding of the concept of zero

Ayo 教材 (動物行為學 2010)
83
Sheba, an amazing chimpanzee
 Using a touch screen, Sheba can indicate the Arabic
numeral that describes a group of objects
 She can add numbers:


If three small groups of objects are put in three separate
places around the room, she can visit them in turn and
then correctly choose the numeral that represents the
sum
If the three groups of objects are replaced with numeral
cards she can still choose the numeral that represents
the correct total
Ayo 教材 (動物行為學 2010)
84
Animals understand concepts
 Alex understood the concepts of same and different
 Brain size does not reliably indicate the ability of animals to do
tasks such as concept learning
 Alex’s brain was the size of a walnut
 Bees can learn to distinguish between same and different
 Pigeons form concepts such as “tree” or “water” or “human”
 Pigeons can identify an example of a particular category, such
as a person
 They recognize water in various forms: a droplet, a river, a
lake
Ayo 教材 (動物行為學 2010)
85
Self-recognition and perspective taking
 The “mark test” examines if an animal recognizes itself
After a mark is placed on its face, the animal is shown a mirror
 If the subject recognizes itself, it sees that it has an odd mark,
and touches and grooms toward the marked area
 Species that have passed the mark test
 Chimpanzees
 Dolphins turn their bodies to inspect marks in the mirror
 Asian elephants touch marks with their trunks

Ayo 教材 (動物行為學 2010)
86
Do animals have a concept of self?
 We can ask whether animals can take the perspective of
other individuals

And understand what others know and do not know
 When given a choice, subordinate chimpanzees
selected a piece of food not visible to the dominant
individual

The subordinate was aware of which piece of food the
dominant could see
Ayo 教材 (動物行為學 2010)
87
The “theory of mind”
 There is a growing literature on “seeing and knowing” and the
attribution of knowledge and mental states
 Many species have been tested with a range of clever
experiments.
 A compelling idea about the evolution of the “theory of mind:” is
that it is driven by social complexity
 The social environment creates new selection pressures for the
evolution of “social intelligence”
 The ability to learn and keep track of relationships among
other individuals may be evolutionarily advantageous
Ayo 教材 (動物行為學 2010)
88
Summary
 Learning: a change in behavior as a result of
experience
 Excluding maturation of the nervous system, fatigue,
or sensory adaptation
 Learning is divided into: habituation, classical
conditioning, operant conditioning, latent and
social learning
 The ability to learn has a genetic basis
 Animals demonstrate cognitive skills in tasks besides
learning
Ayo 教材 (動物行為學 2010)
89
 Tools are objects that an animal uses to reach a goal
 Detour behavior: an animal takes an indirect route to
a goal
 Many animals understand abstract concepts
 Some animals recognize themselves
 Some understand that others do not have the same
knowledge
Ayo 教材 (動物行為學 2010)
90
問題與討論
[email protected]
 Ayo 台南 NUTN 站
http://myweb.nutn.edu.tw/~hycheng/
Ayo 教材 (動物行為學 2010)
91