PJ animal cognition

historical context of animal intelligence

Explores the evolution of methods used to study animal intelligence, from anecdotal observations to controlled experiments.

associative learning

Emphasizes the role of learning processes in intelligent behavior.

pavlovian conditioning

Learning through association between a neutral stimulus and an unconditioned stimulus.

Instrumental Conditioning

Learning through consequences, where behavior is modified by reinforcement or punishment.

discrimination learning

Ability to distinguish between different stimuli and respond accordingly

categorisation

Grouping stimuli based on shared characteristics, facilitating efficient responses

curiosity with studying animal cognition

• Animal behaviour and mental states are inherently interesting.

• Understanding the evolution of cognition and human-animal differences.

• Insights into human-animal interactions.

human welfare with studying animal cognition

• Animal models help study human psychological and neurological disorders (e.g. schizophrenia, addiction, PTSD).

• Enable experiments not possible in humans (e.g. brain lesions, pharmacological injections, histology).

• Better understanding of animals leads to better scientific models of human conditions.

animal welfare with studying animal cognition

• Knowledge of animal cognition informs how we treat animals in various settings.

• Mental states must be inferred from behaviour using observation and experimentation.

• Historical views shaped by figures like Romanes (anthropomorphism) and Lloyd Morgan (parsimony).

• Ongoing debate on approaches—eye-gazers vs. behaviourists vs. experimenters (Fouts, Hayes).

evidence of tool use in animals

• Dolphins use sponges for foraging.

• Chimps use spears, especially adolescent females.

• Animals often use natural objects (stones, sticks) and sometimes modify them (e.g. stripping leaves).

• They may combine tools (e.g. stone hammers) but rarely show evidence of complex modification like shaping stone tools.

Thorndike

cats learn by trial & error in puzzle boxes

Köhler

chimpanzee sultan appeared to solve problems through insight, perhaps based on prior trial-and-error learning

Weir et al. (2002)

a crow bent a wire into a hook, however had prior experience

tube task

Capuchins and apes showed tool use but struggled with more complex tools (e.g. H stick), indicating partial understanding.

trap tube task

Most chimpanzees failed; one succeeded but did not generalise to an inverted version, suggesting a lack of causal understanding

why is understanding of animals & tools important?

• Animals can solve problems simply by learning associations

• However greater understanding allows for insight & complexity

why does categorisation matter?

• Categories help classify new stimuli efficiently.

• Two major theories:

  • Exemplar theory: Compare new items to all previously seen examples.

  • Prototype theory: Compare to an average representation.

examples of category learning in animals

• Pigeons: Can learn abstract categories (e.g., water, people, fish, paintings).

• Capuchins: Discriminate between people and non-people but may rely on low-level cues (e.g., colour).

• Feature-based learning: Most categorisation based on stimulus features rather than abstract concepts.

concrete vs abstract categories

• Concrete: Based on perceptual features (e.g., faces, colours).

• Abstract: Based on relationships (e.g., "son", "same/different").

same/different judgement in animals

• assesses whether animals can understand abstract relational concepts, not just physical features

• assesses whether animals can understand abstract relational concepts, not just physical features

• Pigeons struggle unless given extensive training.

• Alex the parrot: Could verbally answer same/different questions.

• Abstract relational rules (e.g., opposites) are harder but possible.

second-order relationships

• Understanding the relationship between relationships.

• Tested in chimpanzees: Results mixed.

• With support, they can learn relationships between relationships.

what is tested in number cognition

• Relative numerosity (more vs. fewer).

• Absolute number (specific quantities).

• Ordinal scale (ranking).

• Interval scale (equal spacing between numbers).

• Symbol use.

examples of number cognition in animals

• Pigeons: Judged number of red dots → proportion-based decisions.

• Rats: Counted food rewards in a sequence.

• Monkeys: Represented absolute & ordinal numbers.

• Sheba: Used numerals to count food and objects (interval scale).

• Alex: Verbally answered numerical questions.

• Honeybees: Can perform simple addition/subtraction with symbols.

clever Hans effect

• Hans “solved” math by reading human cues.

• Led to caution in interpreting animal abilities → need for controlled testing.

social learning in animals

• Requirements: Behaviour is learned, socially acquired, and persists.

• Social facilitation: Mere presence of others increases behaviour (not true learning).

• Stimulus enhancement: Attention drawn to an object by another's action.

• Examples:

  • Rats: Learn food preferences from others.

  • Monkeys: Observe fear reactions to snakes.

  • Quail: Match demonstrator’s actions (peck vs. tread).

imitation in animals

• True imitation: Copying the same action (not just the result).

• Two-action test: Best evidence (e.g., Japanese quail).

• Dogs showed imitation depending on context (Range et al., 2007).

• Imitation may involve mirror neurons (Heyes, 2010).

self-recognition in animals

• Mirror test: Self-directed behaviour using a mirror = possible self-awareness.

  • Chimpanzees, orangutans pass; most animals fail.

  • Mark test: Touching marks visible only in mirror.

  • Culture affects performance (e.g., Fiji vs. US children).

• Trained pigeons passed mirror test (Epstein et al., 1981).

• Mirror tests show body recognition, not necessarily deep self-awareness.

theory of mind (ToM)

• Definition: Attributing mental states to others.

• Evidence:

  • Deception in wild.

  • Povinelli: Chimps slow to learn attribution of knowledge.

  • Hare et al.: Chimps can act based on what others see—but may be simple discrimination.

animal communication

• Bees: Dance conveys direction & distance of food.

• Vervet monkeys: Alarm calls specific to predators; may be referential but inflexible.

• Communication ≠ Language (limited flexibility, mostly innate).

language research in animals

• Vocal attempts failed (e.g., Viki, Gua).

• Sign/symbol use (e.g., Washoe, Nim, Koko, Kanzi):

  • Can learn symbols.

  • Limited productivity, syntax, and meaningful combinations.

• Best evidence for comprehension:

  • Kanzi: Followed novel instructions, some syntax understanding.

  • Dolphin Akeakamai: Understood word order.

• Most output is instrumental (requests), not generative.