Changing Directions in the Study of Conditioning

Key Learning Goals

• Articulate the theoretical significance of conditioned taste aversion and preparedness.

• Understand the theoretical implications of research on latent learning, signal relations, and response-outcome relations.

Changing Directions in the Study of Conditioning

• Science is constantly evolving, with changes occurring in the study of conditioning due to new research and thinking.

• Two major changes in thinking about conditioning:

  • Recognition that an organism's biological heritage can limit or channel conditioning.

  • Increased appreciation of the role of cognitive processes in conditioning.

Recognising Biological Constraints on Conditioning

• Traditional assumption: The fundamental laws of conditioning have great generality and apply to a wide range of species.

• Until the 1960s, most psychologists assumed associations could be conditioned between any stimulus an organism could register and any response it could make.

• Findings have demonstrated limits to the generality of conditioning principles, imposed by an organism's biological heritage.

Conditioned Taste Aversion

• Martin Seligman's experience with 'sauce béarnaise syndrome':

  • Dined out, enjoyed steak with sauce béarnaise.

  • Six hours later, developed stomach flu and severe nausea.

  • Subsequently, the smell of sauce béarnaise alone nearly made him throw up.

• Conditioned taste aversion is not unique; many people develop aversions to food followed by nausea from illness, alcohol intoxication, or food poisoning.

• Seligman puzzled by his 'sauce béarnaise syndrome' because:

  • It appeared to be a straightforward result of classical conditioning:

  • Neutral stimulus (the sauce) paired with an unconditioned stimulus (the flu), causing an unconditioned response (the nausea).

  • Béarnaise sauce became a conditioned stimulus eliciting nausea.

  • Violated certain basic principles of conditioning:

  • Lengthy delay of six hours between the CS (the sauce) and the US (the stomach flu) should have prevented conditioning from occurring.

  • In laboratory studies, a delay of more than 30 seconds between the CS and US makes it difficult to establish a conditioned response.

  • Conditioning occurred in just one pairing.

  • Only the béarnaise sauce became a CS eliciting nausea; other stimuli present in the restaurant (plates, knives, tablecloths, his wife) did not trigger Seligman's nausea.

• John Garcia and his colleagues solved the riddle of Seligman's sauce béarnaise syndrome:

  • Conducted studies on conditioned taste aversion, manipulating the kinds of stimuli preceding the start of nausea and other harmful experiences in rats.

  • Used radiation to artificially induce the nausea.

  • Found that when taste cues were followed by nausea, rats quickly acquired conditioned taste aversions.

  • Taste cues followed by other types of harmful stimuli (such as shock) did not lead to conditioned taste aversions.

  • Visual and auditory stimuli followed by nausea also failed to produce conditioned aversions.

  • It was almost impossible to create certain associations, whereas taste-nausea associations (and odour-nausea associations) were almost impossible to prevent.

• Theoretical significance of the unique readiness to make connections between taste and nausea:

  • Garcia argues it is a by-product of the evolutionary history of mammals.

  • Animals that consume poisonous foods and survive must learn not to repeat their mistakes.

  • Natural selection will favour organisms who quickly learn what not to eat.

  • Evolution may have biologically programmed some organisms to learn certain types of associations more easily than other types.

Preparedness and Phobias

• Martin Seligman (1971) argues evolution has also programmed organisms to acquire certain fears more readily than others because of preparedness.

• Preparedness involves species-specific predispositions to be conditioned in certain ways and not others.

• Preparedness can explain why certain phobias are vastly more common than others.

• People tend to develop phobias to snakes, spiders, heights, and darkness relatively easily.

• Phobic fears of hammers, knives, hot stoves, and electrical outlets are rare, even after painful experiences with these objects.

• Common phobic objects were once genuine threats to our ancient ancestors.

• A fear response to such objects may have survival value for our species.

• Evolutionary forces gradually wired the human brain to acquire conditioned fears of these stimuli easily and quickly.

• Laboratory simulations of phobic conditioning provide support for the concept of preparedness.

  • Participants shown slides of common phobic stimuli (snakes, spiders, etc.) and neutral stimuli (flowers, mushrooms) or modern fear-relevant stimuli (guns, knives).

  • Common phobic stimuli paired with mild electrical shock.

  • Phobic stimuli produced instant conditioning, a greater fear reaction, and increased resistance to extinction.

• Öhman and Mineka (2001) elaborated on the theory of preparedness, outlining the key elements of an evolved module for fear learning:

  • Preferentially activated by stimuli related to survival threats in evolutionary history.

  • Automatically activated by these stimuli.

  • Relatively resistant to conscious efforts to suppress the resulting fears.

  • Dependent on neural circuitry running through the amygdala.

Recognising Cognitive Processes in Conditioning

• Pavlov, Skinner, and their followers traditionally viewed conditioning as a mechanical process.

• Stimulus-response associations are 'stamped in' by experience.

• Mainstream theories of conditioning did not allocate any role to cognitive processes.

• Research findings have led theorists to shift toward more cognitive explanations of conditioning.

Latent Learning and Cognitive Maps

• Edward C. Tolman (1932, 1938) challenged the conventional view of learning.

• Tolman and his colleagues conducted studies that posed difficult questions for the dominant views of conditioning.

• In one landmark study (Tolman & Honzik, 1930), three groups of food-deprived rats learned to run a complicated maze over a series of once-a-day trials.

  • Group A: Received a food reward when they got to the end of the maze each day; performance improved gradually over 17 days.

  • Group B: Did not receive any food reward; showed only modest improvement in performance.

  • Group C: Did not get any reward for their first ten trials in the maze but were rewarded from the eleventh trial onward; showed little improvement in performance over the first ten trials, but after finding food in the goal box on the eleventh trial, they showed sharp improvement on subsequent trials.

• Tolman concluded that the rats in Group C had been learning about the maze all along but had no motivation to demonstrate this learning until a reward was introduced.

• Tolman called this phenomenon latent learning: learning that is not apparent from behaviour when it first occurs.

• These findings presented a challenge for the current view of learning because:

  • Learning can take place in the absence of reinforcement.

  • The rats who displayed latent learning had formed a cognitive map of the maze (a mental representation of the spatial layout).

• Tolman (1948) conducted other studies that suggested that cognitive processes play a role in conditioning.

• His ideas were ahead of their time and mostly attracted denials and criticism from the influential learning theorists of his era (Hilgard, 1987).

• In the long run, Tolman's ideas succeeded, and models of conditioning were eventually forced to incorporate cognitive factors.

Signal Relations

• Robert Rescorla (1978, 1980) was influential in demonstrating the importance of cognitive factors in classical conditioning.

• Rescorla asserts that environmental stimuli serve as signals and that some stimuli are better, or more dependable, signals than others.

• A 'good' signal is one that allows accurate prediction of the US.

• Manipulated the predictive value of a conditioned stimulus by varying the proportion of trials in which the CS and US are paired.

• In one study, a CS (tone) and US (shock) were paired 100% of the time for one group of rats and only 50% of the time for another group.

• The CS elicited a much stronger fear response in the group that had been exposed to the more dependable signal.

• Studies of signal relations have shown that the predictive value of a CS is an influential factor governing classical conditioning (Rescorla, 1978).

• Classical conditioning may involve information processing rather than reflexive responding.

Response-Outcome Relations and Reinforcement

• Studies of response-outcome relations and reinforcement also highlight the role of cognitive processes in conditioning.

• Example: Studying hard while repeatedly playing a classic Brenda Fassie song the night before an exam. Earning an A on the exam.

  • Does this result strengthen the tendency to play Brenda Fassie music before exams? Probably not.

  • Recognition of the logical relation between the response of studying hard and the reinforcement of a good mark, and only the response of studying will be strengthened (Killeen, 1981).

• B. F. Skinner argued that 'superstitious behaviour' could be established through non-contingent reinforcement.

  • Non-contingent reinforcement occurs when a response is accidentally strengthened by a reinforcer that follows it, even though delivery of the reinforcer was not a result of the response.

  • In a classic study, Skinner (1948) put eight pigeons in operant chambers that were set up to deliver reinforcement every 15 seconds, regardless of what responses the pigeons were making.

  • Six of the eight pigeons started displaying quirky, superstitious responses, such as head-bobbing or turning counter-clockwise.

• Skinner's theory held sway for many years, but theorists eventually turned to more cognitive models of superstitious behaviour.

• Superstitious behaviour is extremely common, and accidental reinforcements may sometimes contribute to these superstitions, along with various types of erroneous reasoning (Vyse, 2014).

• In one study participants were asked to use three golf putters for a laboratory putting task that supposedly was investigating how a lack of visual feedback would affect performance (Churchill, Taylor, & Parkes, 2015).

  • The verbal feedback was manipulated so that it was honest when they used one putter, very negative when they used a second putter, and very positive when they used a third putter.

  • When they were later asked which putter they would like to use in a competition, 22 of 28 participants chose the positive or 'lucky' putter that had been paired with fake reinforcement.

• Research suggests that people are especially prone to superstitious behaviour when they are in situations where their performance will be evaluated (Hamerman & Morewedge, 2015), which probably explains why superstitions are common in sports contexts.

• Athletes exhibit superstitious responses, such as wearing a special pair of socks, eating the same lunch, going through special rituals, and so on, to enhance their chances of success (Bleak & Frederick, 1998).

• South African cricketer Makhaya Ntini carried a piece of cow dung in a plastic bag with him as a good luck charm (Fox Cricket, 2018).

• Athletes appear to rely on superstition more when the level of challenge increases and when there is greater uncertainty about the outcome (Domotor, Ruiz-Barquin, & Szabo, 2016).

• Many people compulsively need to 'knock on wood' after mentioning their good fortune in some area.

• One study (Risen & Gilovich, 2008) showed that many people subscribe to the belief that it is bad luck to 'tempt fate'.

• Contemporary research on superstitious behaviour tends to attribute it to normal cognitive biases and errors that promote irrational reasoning (discussed in Chapter 8) rather than to the unpredictable whims of operant conditioning (Pronin et al., 2006; Wegner & Wheatley, 1999).

• Risen (2016) explains how many normal cognitive tendencies foster superstitious thinking, even in bright, educated and logical individuals.

• Even when people realise that a superstition does not make sense, they sometimes will still allow the irrational belief to influence their feelings and behaviour.

• Reinforcement is not automatic when favourable consequences follow a response.

• People actively reason out the relations between responses and the outcomes that follow.

• When a response is followed by a desirable outcome, the response is more likely to be strengthened if the person thinks that the response caused the outcome.

• Modern, reformulated models of conditioning view it as a matter of detecting the contingencies among environmental events (De Houwer, 2014; Schachtman & Reilly, 2011).