Exclusive book extract by Wulliam Hartston

Journalist and author William Hartston, the writer of the Beachcomber column in the Daily Express and a previous Gogglebox regular, has written 25 books. His latest, Knock Knock, is subtitled ‘in pursuit of a grand unified theory of humour’ and seeks to examine the science, philosophy and history of comedy in all its forms. In this exclusive extract, he investigates some of the neurological research into how the brain reacts to jokes…

Until recently, we have not had much idea what goes on in the brain when we laugh at something. Indeed, we have, until recently, not had much idea of what happens in the brain at all. The task of explaining humour therefore fell mainly to philosophers and psychologists but, as we have seen, they came up with, at best, partial theories and explanations. In recent years, however, that has begun to change as neurologists have increasingly developed techniques to examine the human brain in detail.

Until the late 20th century, most of what we knew about the brain came from observations of people who had suffered injuries to that organ. The functions of different brain regions was deduced from observing the deficiencies resulting from damage to such areas.

The old theory, generally held until around 1980, was linked to the idea that the two hemispheres of the brain were responsible for different activities: the left side was thought to be the domain of language and logic, while the right side was creative and artistic, which included responsibility for humour appreciation. For the most part, that idea is now rejected or at least considered far too simplistic, yet several studies on humour seemed to support it to some extent at least.

In one study of subjects who had suffered localised brain injury, damage to the left hemisphere was found to have little effect in appreciating humour, but damage to the right resulted in unpredictable changes. Another study confirmed that right- hemisphere damaged patients had a greater tendency to choose total non-sequiturs as the funniest captions to cartoons and were liable to be unable to explain the humour, even when presented with the correct caption.

A possible explanation of this confusion was offered in the idea that correct interpretation of a joke can be a two-stage procedure: first, the building up of expectations based on the natural interpretation of the set-up, then the reassessment of everything when the punchline confounds those expectations. A non-sequitur ending can be perceived as a surprise and seen as just as humorous as the correct ending because the damaged right hemisphere cannot resolve the incongruity in either case.

This interpretation may be supported by a case reported in 1998 concerning a patient undergoing brain surgery for epileptic seizures. As was common in such cases, the patient was conscious during the procedure while electrical stimulation was applied to various areas of her brain to locate the precise place where surgery was necessary. When such stimulation was applied to one small area of the left frontal lobe, it consistently provoked genuine laughter. When asked why she was laughing, the patient was reported to have described whatever she was looking at as ‘funny’, even thought there was no humour in it at all.

As with the earlier subjects with right-frontal-lobe damage, she was apparently finding something funny without being able to explain why. Interviewed about her experience later, however, she said her response had nothing to do with what she was looking at but her genuine laughter was caused by her laughing at something that wasn’t funny.

Hang on a moment – that explanation is really weird. If she is laughing because she thinks it is funny that she is laughing at nothing, what caused the laughter that she is laughing at in the first place? She saw something that wasn’t funny and laughed at it, then laughed because she thought it funny that she was laughing.

Once again these relationships between a patient’s behaviour and activation of certain parts of the brain emphasise the difference between humour and laughter: under normal circumstances, the first is an intellectual response to a type of incongruity, while the second is the emotional, physiological response to the intellectual realisation. In the above case, however, it seems that the physiological response of laughter came before the intellect worked out a reason for it.

The trouble with all those experiments is that the sample is highly selective, being restricted either to brain-damaged individuals or people undergoing surgery and is therefore atypical of the general population.

At the start of the 21st Century, however, all that began to change as the science of brain scanning for the first time allowed perfectly normal people to look at cartoons, listen to jokes and even watch funny television programmes or films and make comments on them while having their brains precisely scanned. The development of magnetic resonance imaging techniques in neuroscience then opened up many possibilities to investigate the brain more deeply.

Functional magnetic resonance imaging (fMRI) is a technique that measures brain activity by detecting changes associated with a type of blood flow. This technique is based on the fact that cerebral blood flow and neuronal activation in the brain are closely connected. When a certain brain part is working, the blood flow to that region also starts to increase. We can therefore measure brain activity by detecting changes in blood flow.

The discovery that brain activity correlates with blood flow was made in the 1890 but it took a century before an effective way of using that finding was developed. Quite how this is done need not bother us, but for anyone interested, the basic component of an fMRI scanner is a very powerful electromagnet that can detect the magnetic signals from the hydrogen nuclei in water. Active brain areas need more oxygen, which is delivered by haemoglobin in red blood cells, so the greater the activity in any area, the greater the blood flow and the greater the magnetic signal from the hydrogen nuclei, which is what the fMRI measures and displays in images showing different intensities in various colours.

The first use of fMRI in humour research was a great success, pinpointing the precise areas of the brain that were activated by certain jokes. It was written up in an unpretentious two-page paper by Vinod Goel and Raymond J. Dolan in Nature Neuroscience in 2001 with the title ‘The functional anatomy of humor: segregating cognitive and affective components’.

Explaining their experimental method, the researchers say: ‘We scanned 14 right-handed normal subjects using event-related fMRI while they listened to jokes.’ The term ‘event-related’ simply meant that a subject’s cognitive processes had been separated into discrete points in time (referred to as ‘events’), which allowed differentiation of their associated fMRI signals. In this case, the events consisted of 60 jokes divided equally between ‘phonological jokes’ (based on puns or double meanings) and ‘semantic jokes’ based on understanding and knowledge of the world. The examples they gave in the paper were as follows:

Phonological: Q. Why does the golfer wear two pairs of pants? A. Because he got a hole in one.
Semantic: Q. What do engineers use for birth control? A. Their personalities.

For comparison purposes, the subjects were also told the jokes with changed punchlines that were not funny: for example,  ‘It was a very cold day’ for the golfing joke and ‘The pill’ for the birth-control joke. The brain scans were then examined to see what differences they showed between semantic and phonological jokes or between funny and unfunny ones.

Subjects were also asked to say whether they found each joke funny and to rate it for funniness on a scale of 1 to 5. The results were intriguing and very much confirmed both earlier thoughts about humour and the earliest results of brain scans.

Phonological jokes based on puns or double meanings were shown to result in greater activity in subjects’ left inferior prefrontal cortex, which earlier studies had shown were involved in phonological (sound-related) processing. Semantic jokes, however, produced increased activity in bilateral posterior temporal lobes, which had previously been shown to be active in language processing.

Many jokes, of course, contain both phonological and semantic aspects, but this study was restricted to jokes that were purely one type or the other. For example: ‘Q: What’s orange and sounds like a parrot? A: A carrot’ is purely phonological but ‘Why did Cleopatra bathe in milk? Because she couldn’t find a cow tall enough for a shower!’ is semantic, with nothing phonological about it at all.

The parrot/carrot joke would thus have been appreciated by the left inferior prefrontal cortex of anyone hearing it, while the Cleopatra/cow story would have headed for the bilateral posterior temporal lobes.

The funniest jokes of either type were reported to activate the medial ventral prefrontal cortex, which had previously been seen to be active in people when they are receiving rewards. This last finding, which was confirmed in several later studies, suggested that humour perception, which often takes the form of laughter, is a rewarding experience – something that all of us, with the possible exception of Plato and his humour-sceptic followers, have suspected for thousands of years.

Amusingly, Goel reported that in this study, they deliberately chose only mildly funny jokes, because they did not want their subjects bursting out laughing, which could result in head movements that might interfere with accurate fMRI readings. In general, their results were supported and expanded by later research of a similar nature, including a study in which subjects looked at cartoons and another study involving participants watching episodes of Seinfeld or The Simpsons. In the latter study by William Kelley and Joseph Moran in 2004, the posterior temporal lobes (which had previously been seen to be active in the perception of incongruities) were seen to light up when the subject got the point of the joke, while the amygdala (which has a strong role in controlling emotions) was seen to be active during laughter.

The picture grew still more complex with a study by Watson, Matthews and Allman in 2006 in which subjects looked at Far Side cartoons by Gary Larson while undergoing fMRI scans. Some of the cartoons had captions while others were purely visual and the scans perhaps unsurprisingly showed clear differences between brain activity between the two types: high-level visual areas were activated during visual humour whereas classic language areas were activated in response to language-dependent cartoons with captions. The results also showed that both types activated a common area that included the amygdala and midbrain regions, reflecting the euphoric component of humour.

To throw two more brain areas into the humorous mix, however, they also identified activity in the anterior cingulate cortex and fronto-insular cortex, which are known to have evolved more recently. The paper’s authors conclude: ‘These results suggest that humor may have coevolved with another cognitive specialization of the great apes and humans: the ability to navigate through a shifting and complex social space.’

As if the involvement of all these different brain regions in humour perception was not complicated enough, a piece of research in 2016 added yet another wrinkle when it occurred to two researchers at the University of Southern California to investigate the difference between humour perception and humour generation.

For that purpose, Ori Amir and Irving Biederman, who had already used fMRI scans in previous studies of humour, recruited a number of successful professional comedians, promising amateur stand-up comics and a group of non-comedians, and asked them to write captions to cartoon pictures which had had their original captions and all other verbal material removed. Before each cartoon was presented to them, the subjects were given one of three instructions: to generate a humorous caption; to generate a bland, non- humorous caption; or just to look at the picture without adding any caption.

The subjects were also asked to rate the humour of each caption generated under the first condition and ratings were also obtained from an independent group. The fMRI scans were investigated for differences between brain activity in the three conditions, differences between highly rated captions and low-rated, and differences between the professional, the amateur and the non-comedian groups.

The most intriguing finding was between the professionals and non-professionals. For all groups, the results identified activity in the medial prefrontal cortex and the temporal association regions of the brain when jokes were produced, but the results showed that the more experienced someone was at doing comedy, the more activation there was in the temporal lobe.

This is known to be the region of the brain that mainly functions around hearing and selective listening and is thus vitally important for understanding speech and visual information. It is also the area where abstract information, semantic information and creative associations are all put together in a meaningful way.

It mainly receives sensory information such as signals and speech from the ears to form the basis for our understanding or comprehension. In fact, if we did not have a temporal lobe in our brain, we would not even be able to understand that someone was talking to us. This lobe is special because it interprets all the different types of sounds transmitted from the sensory receptors of the ears.

By contrast, the prefrontal cortex, which the unprofessional groups were relying on, is responsible for planning complex behaviour and decision-making. Funniness ratings were also found to be higher when temporal regions of the brain were used.

Ori Amir summed it up by saying that: ‘The more experience you have doing comedy, the less you need to engage in the top-down control and the more you rely on your spontaneous associations.’

Like so many other mental activities, you learn to do it almost without thinking instead of having to work it out every time. This finally begins to explain why spontaneous humour tends to be funnier than humour that has been laboriously worked out: it comes with experience when our humour production is able to move on to a more creatively efficient part of the brain.

For centuries, philosophers, moralists and psychologists had tried to get to grips with the nature of humour but reached few conclusions. Only when neurologists entered the fray with fMRI scans was progress made, but many questions remain to be answered. The results outlined above identify the parts of the brain we use in performing the activities we must undertake in order to find something funny, such as recognising an incongruity or switching from one interpretation of something to another.

We have also made great strides in discovering which region of the brain is responsible for causing the physiological reaction of laughter and we have confirmed that jokes makes us feel good. Yet we still cannot point to a specific location in the brain where humour resides.

Perhaps ‘humour’ is just a convenient umbrella term to cover a multitude of sensations that have a feel-good effect. The search for the nature of humour may be an illusion, which would, of course, explain why so many diverse theories of humour have been proposed. It could be that all of them are right, setting off different parts of the brain but all ending by provoking laughter.

Published: 10 May 2023