Chomsky's review of Skinner Flashcards
Chomsky takes a closer look at Skinner’s Behaviourist paradigm. What does he say about the applicability of this paradigm to complex (verbal) behaviour?
Skinner has formulated a theory in which verbal behaviour can be explained on the basis of stimulus and reinforcement. His ideas are based on factors (observable in the lab), which were often the result of conditioning experiments with animals. Chomsky believes that Skinner’s findings are too specific and that the laws he assumes cannot be translated into complex verbal behaviour (or only in very broad terms, which makes their usefulness debatable).
Behaviourists explain human behaviour using stimulus-response relationships. According to Chomsky, what dilemma arises when the terms stimulus and response are generalized to complex human behaviour?
Stimulus and response must be defined. This raises the questions: Which physical events count as stimuli? Which part of the behaviour shown is a response? You cannot control and investigate stimulus and response (in terms of language) – is a word a stimulus, or is it a sentence?
If the terms are defined too broadly, anything can be a stimulus and anything can be a response: in that case, you can no longer study legitimate links between stimulus and response. This makes it impossible to come up with an explanation for behaviour.
When we apply too narrowly defined concepts (as Skinner does with ‘Response = pressing a button’ in highly controlled settings), the behaviour can be explained, but it has little meaning because we can only see that specific reaction in that specific laboratory setting as behaviour.
dus: te breed = geen echte associatie, dan kan je het van alles zeggen. te specifiek = niet meer generaliseerbaar
According to Chomsky, what is the problem with Skinner’s definition of reinforcement?
It is difficult to define the stimuli that have a reinforcing effect in real life (because: sometimes silence is reinforcing, sometimes the reinforcer comes centuries later…) and, what has a reinforcing effect can vary between and within people over time).
Skinner argues that many stimuli are reinforcers, but Chomsky thinks that this makes it too vague.
dus lastig om te bepalen wat stimuli zijn, en wat responses zijn -> maakt het ook weer lastig wat precies reinforcement is.
b) Chomsky not only sees issues with Skinner’s definition of reinforcement, but also refutes its importance in learning. How does he do this?
Chomsky quotes a number of experiments that serve as examples of reinforcement not always being necessary for learning. He uses an example of a rat that displays learning behaviour purely out of curiosity (instead of being rewarded).
According to Chomsky, what are important factors in language learning? Describe these and then formulate them in a broad sense: what do Chomsky’s ideas boil down to and how does that core principle differ from Skinner’s?
Reinforcement, observation, curiosity (inquisitiveness) and the tendency to imitate. The child uses the ability to generalise, hypothesise and process information. This requires a system that is innate or that is developed as the child grows older (genetically determined maturation).
Examples
- Children learn behaviour and language by observing and imitating others. According to Skinner, the parents provide reinforcers. But Chomsky argues that in many families the parents do not pay so much attention to language. Immigrant children often learn a second language in the streets by observing other children.
- If parents make no effort to teach certain words, children will still be able to repeat them.
- Children can create and understand new words that fit within the rules of the language. The behaviourist argues that this is due to stimulus generalisation. But that is just another name and does not explain anything according to Chomsky.
These examples show that there is a fundamental process (Chomsky) that is independent of the feedback from the environment (Skinner)
Skinner believes that if a speaker expresses himself, there must be an objective link to the environment that can be described. For example, someone who asks a question will have a lack of something (deprivation). In that case, that lack of something is the reason for asking the question. What does Chomsky find problematic about this definition?
Chomsky finds this problematic, because it would imply that people are constantly experiencing a lack of all kinds of things. He also argues that some deprivations are difficult to describe (such as “I’ll fix this for you.” What is lacking here?).
Chomsky uses the example of a fox to address the concept of a tact. What point is Chomsky trying to make with this example?
A tact is a verbal operant “in which a response of given form is evoked (or at least strengthened) by a particular object or event or property of an object or event.”
a tact is verbal behaviour, which is aimed at ‘improving’ one’s own situation (and is therefore tactful). Parents are described as teaching their child a language for their own benefit, arguing that a verbal child would be ‘more convenient’ for the parent than a non-verbal child.
With the example of the fox, Chomsky explains that person A says “fox” to B. B reacts ‘appropriately’ by looking around, running away and aiming his rifle. Chomsky finds explaining this behaviour complicated. In his example, we assume that B would like to see a fox. Looking around, etc., would result in person B seeing a fox. Is this because this behaviour by B has been reinforced in the past by indeed seeing a fox? That is not necessarily the case. Perhaps he doesn’t even want to see a fox, but still shows the right behaviour.
Whether or not B had seen a fox in the past, or whether or not he wanted to see a fox at that time: B’s ‘appropriate’ behaviour could have been possible in all scenarios.
How does Chomsky view the role of echoic operants in language acquisition?
According to Chomsky, they are not necessary. Echoic operants are responses that generate a sound pattern similar to that of the stimulus (e.g. a child copying what a parent says). According to Skinner, children only do what their verbal community wants them to do, so how correctly the child pronounces something depends on how accurate the feedback from the community is. However, there is no evidence that children only learn phonemes effectively when the community requires ‘perfect correspondence’.
wannt een vader kan bijvoorbeeld heel zwaar praten, maar dan praat het kind niet gelijk zo.
According to Skinner, there is a link between autoclitics and syntax. What are Chomsky’s views on syntax?
Skinner states that autoclitics reflect relationships between phrases and thus form the basis of syntax. Chomsky recognises a deeper effect of grammar and syntax. He offers examples showing that sentences may have the same structure, but differ internally (e.g. in meaning and subtle structures). In other words, he argues that you can’t construct sentences according to standard ‘fill-in model’.
chomsky’s conclusion:
Chomsky concludes by saying that we can’t find out anything about the cause of verbal behaviour until we know enough about language itself. Chomsky recommends first understanding the underlying structures responsible for grammatical understanding. Grammar is a complex and abstract system, which children nevertheless learn quickly, similarly, and often independently of IQ. We can now only derive the functioning of the grammatical system from what manifests: verbal behaviour.
