Savage Rumbaugh ppt

Savage-Rumbaugh et al. (1986) Spontaneous symbol acquisition and communicative use by pygmy chimpanzees (Pan paniscus)

To report on the language acquisition of Kanzi, a bonobo chimp (pygmy chimpanzee), and his younger sister Mulika. The report covers a 17-month period, beginning when Kanzi was aged 2 ½ years. To compare language acquisition in the bonobos with that of two common chimpanzees, Sherman and Austin, studied earlier (1975-1980). Kanzi – a bonobo chimp ( Pan paniscus ) Sherman & Austin – common chimps ( Pan troglodytes ) AIM:

Kanzi – born in captivity in US (1980) His mother was Matata , a wild- captured bonobo, who was the actual intended subject for the language study Matata was not Kanzi’s biological mother. She had kidnapped him from his mother, Lorel , when he was just a few hours old In 1983, Matata gave birth to Mulika , the second bonobo subject in this research She grew up as Kanzi’s sister PARTICIPANTS / SUBJECTS:

A CASE STUDY was carried out, focusing on one bonobo, Kanzi, but also including early data of his younger sister Mulika’s language acquisition. Data gathered by OBSERVATION . ( quantitative ) The design was LONGITUDINAL , reporting over a 17-month period of Kanzi’s language acquisition. (10 years in total) Think! What are the potential STRENGTHS and WEAKNESSES of the method chosen? Can findings from a case study be generalised? - Are findings gained from observation reliable and valid? - Can all variables be controlled in a longitudinal study? METHOD:

The procedure is in 6 parts 1. Description of the environment where the apes were living Naturalistic outdoor environment Indoor environment 55 acre forest Food placed in 17 locations – Kanzi + carers spent most of day collecting it Kanzi communicated about food in the forest, and in a backpack, with lexigrams Mulika also did this as she got older In the day, Kanzi and Mulika helped with domestic chores e.g. food preparation Helped spontaneously with simple jobs like washing up Watched videos of people/animals they knew from the centre Played with toys Played games e.g. tickle, chase… PROCEDURE:

2. The communication system used with the apes The lexigram system: A visual symbol system A keyboard of symbols, lit up when pressed To start with Kanzi’s board had 8 lexigrams By the end of the study, had 256 lexigrams Connected to an electronic voice synthesiser – Kanzi understood spoken English Indoors, the keyboards weren’t mobile – attached to computer Outdoors, Kanzi used portable keyboards and laminated pointing boards Other means of communication: Spoken English Gestures, including 100 ASL gestures

3. The exposure of Kanzi and Mulika to lexigrams – how they learned The lexigram system Kansi was exposed to use of symbols, gestures, and human speech from age 6 months This happened as he watched his mother, Matala, and her keepers No-one tried to train Kanzi directly Mulika learned from observing Kanzi They used the same lexigram system as Sherman and Austin However, Sherman and Austin were taught to use it – Kanzi and Mulika weren’t. They spontaneously used it after observing others Other differences between Kanzi and Mulika, and Sherman and Austin: On Sherman and Austin’s keyboards there was no voice synthesiser – they didn’t understand human speech Sherman and Austin didn’t use lexigrams outdoors – they couldn’t use them if they didn’t light up

Apparatus: Lexigram board – made up of symbols used to stand for words; these symbols brighten up when touched (256 keys). Note : each lexigram was arbitrary i.e. it did not look like the object or word it stood for.

4. Data recording and classifying Classifying the utterances: INDOORS – lexigram use could be automatically recorded OUTDOORS – a record was made by hand and entered into computer later There was a complete record of Kanzi’s utterances from 30-47 months of age For Mulika, the record ran from 11-21 months The utterances were classified as: ‘ correct’ or ‘incorrect’ ‘ spontaneous’ if there was no prompting or cueing ‘ imitated’ if they included a companion’s utterance ‘ structured’ if they were in response to a question Structured questions were used to see if the chimps could give a specific answer Recording:

5. Vocabulary acquisition criteria Establishing reliability: (CONTROLS) Utterance happened spontaneously and was appropriate to what primate was trying to achieve, and could be verified on 9 out of 10 times e.g. if Kanzi indicated he wanted to go to the treehouse it would be verified if he took the experimenter to this location. This was counted as a ‘positive concordance score’ A 4½-hour block of real-time observations were compared with analysis of a video of the primates that was made at the same time. There was 100% agreement on the lexigrams used and their correctness There was disagreement on whether one utterance was spontaneous The video observer counted 9 extra utterances What counted as ‘acquired’?

6. Formal testing of productive and receptive capacities At the end of the study, Kanzi and Mulika were formally tested on all the words in their vocabulary. Kanzi was tested in the 17 th month of the study, aged 46-47 months Mulika was tested aged 18-21 months When? Why? To ensure that their performance was not due to the experimenters giving them cues, or inadvertent glances. This acted as another control measure

How? The Tests. They were shown photographs and asked to select the right lexigram They listened to a word or a synthesised version of the word and then asked to select the correct photograph or lexigram These were ‘blind’ tests and the order of photographs / words was carefully controlled so that the experimenter had no prior knowledge that could bias primate responses.

PROCEDURE EVALUATED bonobos lived at the Language Research Centre (had human companions who also used ASL and spoken English) exposed to lexigram keyboard – which eventually consisted of 256 keys Kanzi & Mulika learned through observation (mother was Matata) – no formal training provided to chimps all utterances were recorded and coded: - correct / incorrect - spontaneous / imitated / structured criteria for learning of symbol: - behavioural concordance measure (what Kanzi said had to match up with what he did) Tests consisted of: - ‘blind’ test in forest (with person unconnected with training) - matching photo to lexigram symbol - matching photo to spoken English - matching lexigram symbol to spoken English

Results 1: Greater use of specific and untutored gestures by K & M Early referential use of lexigrams (M at 12mths) Symbol acquisition: K = 44; M = 37 lexigrams ‘ Blind test’ in forest – K able to lead experimenter out of forest by correctly using photos and lexigrams

Results 2: 80% of K’s utterances were spontaneous Use of combinations (multiple symbols) – 2,500+ correct and mostly about initiating games D A B F

Results 3: Formal Vocabulary Tests Bonobos can understand spoken English whereas common chimps cannot “ Which one is the tomato Kanzi, show me tomato” Primate Number correct / total Matching symbol to English Matching photo to English Matching photo to symbol K 65/66 56/59 55/59 M 41/42 36/41 41/42 A Not tested 3/30 30/30 S Not tested 2/30 30/30

Conclusions K and M (bonobos / pygmy chimps) were different from common chimps. Their results were significantly better compared to the first research subjects. They were spoken to normally which allowed them to learn language and respond without contextual cues. This was suggested to be down to the lack of training and the power of nurture and Cultural learning . 2. The beginning of syntax had been learned by M and K. (rules of sentences). They were able to direct others (request a to act on b) 3. It appears that language of some sort can be learned without being hard wired (agrees with argument that language is learned rather than being innate)

Comparing Pygmy Chimps with common chimps The researchers were able to summaries 4 main differences between the Pygmy chimpanzees and the Common chimpanzees. Kanzi and Mulika comprehended the lexigrams with far more ease, and used then far more spontaneously without the need for training, than Austin and Sherman. Kanzi and Mulika were far more able to comprehend spoken English words. Kanzi and Mulika used lexigrams far more specifically e.g. to differentiate between coke and juice, than Austin and Sherman who used broader categories e.g. food. Kanzi was able to refer to requests involving others. Austin and Sherman never formed requests in which someone other than themselves was the beneficiary.

Evaluating the study Ethical considerations: Chimpanzees being reared away from their natural environment for human research Unable to return to natural habitat Develop human dependency Often frustrated and bored by constant human intervention Skills learnt that are not required in the wild

Ecological Validity- the degree to which the behaviours observed and recorded in a study reflect the behaviours that actually occur in natural settings. EV Strengths found in the study: Kanzi and Mulika were free to roam round centre and in 55 acres of woodland made it more natural setting. They were not instrumentally conditioned but learned by observation as children do. They could also visit their mother Matata EV Weaknesses: Wild animal reared in a human setting for the purposes of research From a young age animals were constantly exposed to human behaviour and nurtured by humans alongside their mother.

What are the strengths and weaknesses of the study? Subjects Method (quantitative) Procedure Ethics

Strengths of the method Improvement of previous studies on language acquisition of primates. Data was gathered under rigorous controls which meant the data was more reliable and valid as less open to bias and subjectivity. Longitudinal study – allows in depth data and shows development over time Ecological validity – Kanzi was observed by researchers whilst outside and roaming freely.

Weaknesses of the method Ethics – animals are vulnerable and cannot express desires like humans. It was clear the subjects became ‘humanised’ – wanted constant human attention. Chimps engaged in human past times like watching TV and drinking coke and eating sweets – not normal in the wild. Difficult for subjects to return to the wild Differences of rearing for K and M compared to S and A.

Participants Small number of subjects. Used a different species of chimp for this study. Differences of rearing for K and M compared to S and A.

Type of data collected A vast amount of data collected – allows for easy comparisons with previous subjects. However, validity of data relies on the validity of the tests, which do not allow for comments on and observations of behaviours that may have occurred during the tests (REDUCTIONIST – means data does not allow for more detailed evidence to be recorded)

How useful was this research? Chimps displayed many characteristics of a human child learning language – this helps us to develop our understanding of how humans learn language. Shows the potential for animals learning language – may have implications for the way we live alongside animals in the future.

Describe ONE change you would make to this research What? How? Why? What would you expect to happen?

Savage Rumbaugh ppt

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Savage Rumbaugh ppt (20)

Savage Rumbaugh ppt