Animal Cognition Flashcards
What are the 8 key components of language?
- Communication
- Voluntary (not just involuntary expressions of emotion)
- Structured - discrete structures within an organised system (phonemes, syllables, morphology, & grammar)
- Symbolic (not just indexes or icons)
- Open-ended (discrete infinity)
- Learnt, culturally transmitted, local dialects & languages
- Displacement - languages can be used to communicate about things that are not in the immediate vicinity
- Conversational - turn taking
Define index, icons, and signs in terms of de Sassure’s (1916) “relationship btwn the signifier and the signified”
Index - signifier is in some way physically or causally linked to the signified. E.g., smoke is linked to fire
Icons - signifier bears a physical resemblance to the signified. E.g., a road sign conveys something non-arbitrary about something else, it’s iconic of the signified
Signs - the relationship btwn the signifier and the signified is arbitrary and purely conventional (symbolic). E.g., the arbitrary relationship btwn a word and what it signifies.
What has been argued to be a key feature that powers human communication? An example of this from research?
Arbitrary symbol use (signs) corresponding to meanings. E.g., global phenomenon of bouba/kiki association with round/spiky.
Explain the claim that vervet monkeys use language, and a critical counter-claim. How would we test these claims? Is it language?
Claim - Monkeys made differing calls for different kinds of threats, for example, a snake vs. an aerial threat. Because there was no causal/physical link btwn the calls and the stimuli, this was suggested to be a demonstration of language in nonhumans.
Critique - Their communication lacked discrete infinity, i.e., if they are capable of complex language, why is its usage limited to a small number of stimuli? Why wouldn’t this complex communication ‘explode’, if they can arbitrarily relate referents, sound, and meaning.
Explain von Frisch’s claim that honeybees use language, and a critical counter-claim. How would we test these claims? Is the dance, language?
Claim - Worker bees perform a waggle dance in an 8 pattern on returning to the hive. Von Frisch found that bees were communicating direction of food - the axis of orientation of the central part of the dance pointed to source of food in relation to the sun.
Could also communicate distance - relation btwn number of times figure 8 circle was danced and distance of food from hive. So could communicate about a potentially infinite number of places, and their distance.
Critique - Actually, information conveyed via a different mode - olfactory cues. Bees return with pollen on them, other bees detect olfactory compounds and detect similar cues when they leave the hive.
How to test - Waggle dance or odour cues, or a combination, conveying info? - German researchers used a mechanical bee that could be controlled - this demonstrated that it was the waggle dance conveying information.
Summary - language? System of communication about things in the world in the here-and-now. Not arbitrary - 1-1 angle of waggle and angle of sun - like pointing. Limited to objects and locations in the world - iconic. Doesn’t have grammar.
Explain Hailman & Ficken’s paper about chickadees having grammar, and a critical counter-claim.
Findings - Claimed birds made 3479 calls, with 4 elements and 362 different sequences, only 11 didn’t obey the rules. They claimed that while there could be any number of repeats of a component (like “chick”) but a ‘grammatical rule’ constrained possible orders.
Counter - Only a finite state grammar- Minimally grammarlike in that there seems to be a Markovian rule (state at time 1 determines state at time 2). BUT, in human utterance, we typically find much more complex grammar than this finite kind.
AND - and only a signal system – not symbolic like human grammar.
Why does Pinker think we care about ape language cognition?
Suggests that if we find there is a bridge between apes and humans, we may take a more optimistic view of human futures - based on their apparent high cooperation and “ethical” like behaviours.
E.g. Great Ape Project said we have a DNA in common, it’s been ‘proven’ they have language, thus we should ban experimentation and afford them legal rights.
What’s the problem with teaching chimps human vocal language? What other method could be used to teach chimps language?
Chimps’ vocal tracts fundamentally different - larynx positioned much higher than in humans - lack ability to make vocalisations found in human language. Mechanical problem. Also - many chimp vocalisations are involuntary, under emotional control. Alternatively, many tried using sign language to teach chimps language.
Discuss claims and criticism of Washoe the chimp’s language abilities
Claims:
- 1960s, Common chimpraised by the Gardeners
- American Sign Language
- 125 words (350 according to wikipedia)
- Allegedly, on seeing a swan for the first time, novel terms coined - “waterbird”
- Media at the time - “SOS from outer space” - highly sentient life on Earth etc.
Criticisms: Didn’t really learn ASL.
1. Only respond, not spontaneous use
2. Interpretation of signs disputed by actual ASL users
3. Jane Goodall, all gestures familiar from prior work with chimps - Washoe hadn’t learnt arbitrary symbols
4. Data not available for scrutiny
5. 25, not 125 words
6. Water Bird, not Waterbird
Discuss claims and criticism of Koko the gorilla’s language abilities. How could we try verify the claims?
Claim:
- Penny Paterson claimed Koko learnt 375 signs
- Understand 2000 English words
- Lie, swear, pun, report dreams
- Gossip in ASL
- IQ 84 – 95!
- Grieved for Robin Williams
Counters:
- ‘Clever Hans’ effect - Koko could be getting cues from human handler
- Was Koko understanding Penny’s communications or just acting on prior positive reinforcement? Could present novel kinds of communications to probe at understanding.
- Independent people to verify the signs - often Koko would give ambiguous responses but these were over-interpreted as deliberate lies/jokes from Koko.
Discuss experiments with Nim Chimpsky and Herb Terrace. How could we try verify the claims? Pinker’s comment?
Claim:
- Herb Terrace aimed for rigorous evaluation of teaching Nim signs, avoiding over-interpretation, Clever Hans effects seen in prior experiments
Result:
1. Just drill - repeated reinforcement
2. No grammar
3. Action/object - e.g., “Nim eat Nim eat”; “Me gum me gum”; “Banana me me me eat”
- Terrace: “Apes can learn many isolated symbols (as can dogs, horses, and other nonhuman species), but they show no unequivocal mastering of the conversational, semantic, or syntactic organisation of language.”
- Hasn’t generalised on the notion of a symbol/arbitrary meanings for referents
Pinker contrasted Nim’s utterances with 3yo already complex grammar - e.g. We going turn light off so you can’t see”
Discuss experiments with Kanzi the bonobo, and what Savage-Rumbaugh did to eliminate methodological issues from previous work
- Careful testing not to accidentally cue Kanzi – e.g. wore mask to conceal eye gestures etc.
- taught Kanzi to recognise truly arbitrary symbols from a lexigram, coupled with speech synthesiser
- Kanzi did well, learnt 100 words (arbitrary symbols)
- No conversation - typically action/object and reactive, not grasp functionality of language
- More convincing but still limited
Explain recursion and its two types
Recursion = the ability to place a grammatical element inside an element of the same kind - an operation that calls back on itself.
often used to create expressions that modify meaning of one of the elements of the sentence. E.g., to take the word nails and give it a more specific meaning, we could use an object relative clause such as “that Dan bought”
Tail recursion = Simple recursion, just adding elements.
E.g.: This is the house that Jack built.
This is the malt that lay in the house that Jack built.
This is the rat that are the malt that lay in the house that Jack built.
True recursion = i.e. centre embedding. The more complex version of recursion that enables human language to be generative.
The malt that the rat ate lay in the house that Jack built.
The malt that the rat that the cat killed ate lay in the house that Jack built.
Explain Fitch & Hauser’s claim that starlings demonstrate recursion, and a counter to the claim from Corballis.
- Trained starlings to learn ‘a’ rattle sound, and ‘b’ warble sound
- trained to discriminate sentences that were grammatical or not - 10k - 50k trials
- two kinds of structures: abab vs aabb (tail recursion); abababab vs aaabbb (true recursion)
- claimed 9/11 birds could parse complex recursion & thus showed ability as complex as uniquely human ability
Counter: (Corballis)
- could be solved (as a starling) by just counting how many of each sound element in a trial - number of ‘a’s and ‘b’s - after many thousands of trials this is feasible
What is a key difference between what Kanzi can do, and what humans can do in terms of language use?
- Complex recursive grammar/syntax
- Creative use
