Our lab communicating our science to you.
We are recruiting for a study with people who stammer - please see here for more information!
It’s been a great year for neuroscientific investigations of domestic dogs. A study in April compared brain responses in dogs and humansusing fMRI (requiring the dogs to lie still I the scanner, to command). This showed that humans and dogs both activate the superior temporal sulcus when they hear emotional vocalizations from dogs and humans, though both species showed an enhanced neural response to calls from their own species. A study hasjust been published using human speech, showing that dogs will preferentially turn to present their right ear to the sounds if they are presented with speech with emphasized segmental (phoneme) information, and with their left ear if the speech has exaggerated pitch (intonation). This is interpreted as indicating that, as in humans, dogs process speech preferentially with the left side of the brain, and intonation in the left (there's an example here).
This kind of study – looking at head turning preferences as an index of hemispheric lateralization – is made possible because of the ways that nervous systems decussate – cross from one side of the body to the other. For example, when I look at the world, the information from the left hand side of both of my eyes (the left visual field) is all sent to the right side of the brain, and vice versa. Likewise in motor control, the left side of my brain controls my right hand and vice versa. Unlike the visual system, the incoming information from the auditory system does not completely decussate, but there is enough dominance of this left to right and right to left pathways that it can be used to find hemispheric asymmetries. Humans are better at listening to speech if it is presented to the right ear, for example, indicating that (for the majority of people) the speech is preferentially processed in the left side of the brain, a finding that is consistent with what we know from patient studies, functional imaging, and studies of which ear people like to hold a phone to. Head turning paradigms are exploiting the same relationship – the idea is that the participant will turn the preferred ear to the sound. So turning the right ear to speech indicates that there is preferential processing of speech in the left hemisphere.
The head turning paradigm is useful as it lets us look at lateralization in animals –such as dogs – whom we can’t ask to make overt responses (e.g. an test for an ear advantage), and this has formed an important tool in studies ofnon human hemispheric specializations.
It’s always good to test the extent that we can assume paradigms are testing what we think they are testing, and a study a few years ago investigated head turning preferences in humans. Their logic was that this is often used as a way of inferring hemispheric asymmetries, and given we know that humans do have lateralized brain responses, do they show a head turning preference? They went out and tested over two hundred German participants with both speech and non speech sounds, and they found that there was no evidence for a right ear dominance when people turned to listen to speech sounds. Indeed, there was a significant tendency to turn the left ear to all the sounds, suggesting that there is a right hemisphere dominance for listening to sounds, generally. Notably, another study in which people (I'm paraphrasing) went round noisy night clubs and muttered at people reported that participants most frequently presented their right ears in an attend to clarify what was being said. But this is not exactly the same as the straightforward head turn paradigm.
This doesn’t mean that the recent dog study isn’t revealing a truly lateralized response – but it does mean that we need to bear in mind that head turn preferences can reflect something more complex.
We have had three wonderful work experience students in our lab for the past two weeks, and they've written a great piece on the work they've been doing for some of our laughter research:
Figure 1- Participants’ answers to the question ‘what makes you laugh?’ The bigger the word, the more frequently mentioned it was.
When we secured a work experience placement at UCL, we expected to receive menial tasks, for example having to make tea or coffee and photocopying. Yet on the very first day of our placement we were handed 2 carrier bags, full of pieces of paper with the word poo (alongside creative doodles) on them. This was the data collected by the laughter lab, who have been investigating over the last few years what made people laugh.
The results were gathered in July 2012 and March 2013, which essentially gave us a snapshot of that period. The high frequency of Miranda [Hart] could possibly be due to the peaked interest in 2012 after the 3rd series of her self-titled show was released . Similarly, the Harlem Shake  went viral in early February 2013, which could explain why quite a few people submitted the dance meme as a suggestion. Another unexpected answer was Mr Bean; which we initially thought was due to the young age of our demographic. We later realised that in July 2012, he performed in the Opening Ceremony of the London 2012 Olympics , viewed by 27 million UK viewers. It is fair to assume that our results are somewhat dependent on popular culture events of the time.
Figure 2- Percentage of categories once the answers had been classified.
Our transcriptions were put into various categories depending on what they entailed. Each transcription was tagged with a maximum of 3 key words. For example ‘Dad falling over’ was categorised as ‘friends/family’ and ‘schadenfreude’. Others were harder to categorise such as people just writing “everything” – we classed these as anomalous results and ignored them, excluding them from our final results. After our initial classification, we refined our data by reducing the amount of categories we had. For example we originally had ‘injury’ and ‘violence’ as two separate categories and we later decided to merge them into ‘schadenfreude’.
We have sorted through data ranging from ‘flowers and snowflakes under a microscope’ to ‘Sachin’s coconut head’. These suggestions and others lacked sense and so were harder to classify; we assumed they were private jokes thus tagging them as ‘jokes’ and a number were censored due to inappropriate content. Several participants also wrote down what they liked rather than what triggered laughter within them i.e ‘I love One Direction’. A handful of answers referred to other answers such as an arrow pointing upwards or ‘this’. Due to not having the original layout of the boards, these answers could not be used in our final data.
If you look at figure 3, common answers included celebrities such as actors, Youtubers and comedians. 38% of celebrities mentioned were comedians, with the jolly Miranda Hart dominating this category. One of the most interesting findings was the significant number of people whose laughter was provoked by social interaction. For example it is quite clear from the figure 1 that ‘friends’ and ‘people’ appeared as the majority. This is further reinforced by the high percentage (24%) of answers that were categorised into ‘friends/family’.
Lucy Kent, Haafizah Khodabocus and George Ponniah Work experience students from Graveney School
 Miranda Hart http://www.youtube.com/watch?v=HFUj5GqTArU
 Harlem Shake http://www.youtube.com/watch?v=iGCy2zp4XuA
 Mr Bean Olympic Cameo http://www.youtube.com/watch?v=BSg_AvqF0qU
Check out Sophie's piece about the difficulty of studying laughter, featured on the Guardian website!
Videos from the article:
Here are some recent (and not-so-recent) videos of Sophie Scott's talks and stand-up comedy routines about Laughter!
2010 UCL mini-lecture about the neuroscience of laughter
2012 TEDxImperialCollege talk
2012 UCL mini-lecture about learning to laugh
2012 Sophie on James May's "Things You Need to Know - The Brain" (S02E02)
2013 Sophie on "Some Boffins with Jokes" - BBC Four
2013 UCL Lunch Hour Lecture - LOLZ! The science of laughter
here is the Random Acts film about our work with Reeps One
the behind the scenes footage is here
the film was made by I Owe Youth which makes the Speech Communication Lab happy as we like puns.
We recently helped with the filming of a Channel 4 Random Acts film with beatboxer Reeps One - there is a shot behind the scenes film here. We'll post a link to the final film shortly!
This study was an online experiment where people (comedians, actors and a control group) were invited to complete a questionnaire testing 4 different kinds of differences in experiences which relate to aspects of psychosis, in the sense that they may map onto experiential aspects of how people’s thoughts and moods may vary (i.e. it’s not a test of psychosis). The idea is to look at comedians as creative people tend to report higher scores on these measures.
(from Ando et al 2014)
The four scales are:
“(a) Unusual Experiences (UnEx), measuring magical thinking, belief in telepathy and other paranormal events, and a tendency to experience perceptual aberrations;
(b) Cognitive Disorganisation (CogDis), measuring distractibility and difficulty in focusing thoughts;
(c) Introvertive Anhedonia (IntAn), measuring a reduced ability to feel social and physical pleasure, including an avoidance of intimacy;
(d) Impulsive Non-conformity (ImpNon), measuring a tendency towards impulsive, antisocial behaviour, often suggesting a lack of mood-related self-control.”
A higher score on this is associated with a higher frequency or stronger experience of the feelings – so a high score on the Introvertive Anhedonia is associated with more feelings of (e.g.) not enjoying social pleasure. There is some evidence of sex differences on these scales:
(From Mason and Claridge, 2006)
From this table, women score more highly on the IntAn and ImpNon scales and men on the CogDis scales. A reason perhaps to try and match for the numbers of men and women, or at least control them across the groups?
In the current paper they note that gender is significantly affecting the experience scores, but as it does not interact with the group variable (actor, comedian or control) when they do ANOVAS for each sub scale, they do not consider it further. However as much of the difference is in the profile of responses across the different O-LIFE subscales, perhaps they should investigate it further. Certainly when they compares across the subscales (just for actors and comedians), it does look like there is significant variation with sex – with the male comedians looking more like the male and female actors, except for the IntAn score, where they’re scoring higher (note this is also different from the 2006 study, where women score more highly on this sub score).
So by this study, the comedians and actors are somewhat different from the control group, except for the IntAn score, where the comedians score more highly than both. However within the actors and comedians, the male comedians look more like the female actors (except for IntAn) and the female comedians look more like the outliers (on the UnEx CogDis and ImpNon subscales). There is something going on here – but it looks like we also need to account for the potential sex differences – or maybe just get more female comedians into the study - before we decide that comedians are unusual. As the authors say, we also need to think more about what actors and comedians do, especially as they look so similar in many ways.
Note – I think they’re using the z scores because the subscales differ in basic distribution of scores.
1. The participants are aged between 8-22yrs, and the analysis splits them into six groups (male and female groups, and three different age groups). Nonetheless, a parametric factor of age would seem absolutely essential and means the movement issues pointed out by @practicalfMRI are a real issue. This also seem problematic as the authors themselves suggest that here are sex-related differences in brain development (effects which they don’t find, I think, see below)
2. The connectivity is base on parcellation of 68 cortical and 27 subcortical regions per subject, and these are used as the nodes for the DTI analysis. This is a very coarse grained analysis of the human brain and assumes that this level and anatomical structure of parcellation is the correct one to address brain connectivity.
3. I can’t find any analysis of the effects which are found across both men and women, either as two groups or in terms of main age effects – i.e. results are described in terms of sex differences, not the regression results. This means that it’s hard to judge the meaning of the results – sex is important, but we know age is important, too. And other studies have claimed that age is even more important than sex. NB there is no age-by-sex interaction in the connection based analysis, which is surprising.
4. What does it mean to say that the male brain is optomised for interhemispheric processing? There are great big structures in the brain which connect the left and right hemispheres – are we to assume that these are non-functional in males, like nipples?
5. You might also assume that we’d get a hint of such processing differences in the clinical literature, but other than some diseases affecting men and women differently in terms of numbers, the effects of stroke, Parkinson’s disease etc. are the same for men and women.