Speaking out

Our lab communicating our science to you.



Help! We're calling all guitarists and beatboxers

posted 20 Apr 2015, 04:32 by Saloni Krishnan   [ updated 20 Apr 2015, 04:39 ]

We're the Speech Communication Lab at the Institute of Cognitive Neuroscience, UCL and we're interested in how mastering a complex skill (like playing the guitar or beatboxing) changes the way brains respond to sound. Other groups have shown that dancers's brains react differently to non-dancers when they see dance. However, does the same apply to different forms of music? We're eager to find out! 

So, if you're a professional guitarist or beatboxer based around London, we'd very much like to hear from you! Feel free to get in touch using the form below, reach out to us on Twitter, or email us. This project is being run by Dr. Saloni Krishnan (@salonikrishnan, s.krishnan@ucl.ac.uk) and Prof. Sophie Scott (@sophiescott, sophie.scott@ucl.ac.uk).


To design this project, we've collaborated with two incredible musicians, Reeps One and Darren LovedayYou can see some of Sophie's previous work with Reeps One here: 

Reeps One x Sophie Scott


If you decide to help us and meet our safety criteria, we'll invite you to come and be scanned at our neuroimaging centre (Birkbeck-UCL Centre for Neuroimaging, 26 Bedford Way, London). The session takes about 2-3 hours. You'd be in the scanner for about an hour, during which time we play you some music and other sounds. We can pay reasonable expenses for your time (and maybe even give you a picture of your brain!). 



people who stammer - we need you!

posted 9 Feb 2015, 01:54 by Sophie Scott

We are recruiting for a study with people who stammer - please see here for more information!

musicians and non musicians and masked speech

posted 23 Jan 2015, 14:59 by Sophie Scott

Download our new JASA paper here. We compared musicians and non musicians on a range of tasks, and unlike some previous work, found no signs of enhanced performance by musicians when listening to speech against other sounds.

turning heads with dogs and humans

posted 15 Dec 2014, 08:14 by Sophie Scott   [ updated 15 Dec 2014, 08:28 ]

some dogs (image from animalia-life.com)

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. 

What makes YOU laugh?

posted 11 Jul 2014, 07:31 by Dana Boebinger   [ updated 11 Jul 2014, 07:51 ]

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 [1].  Similarly, the Harlem Shake [2] 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 [3], 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’. 


Many, it seemed, gained pleasure from seeing people fall over. In fact 75% of answers classed as ‘schadenfreude’ was along the lines of witnessing someone fall over in public. The only real deviation was ‘baby horse falling over.’ Out of all these, only one explicitly said falling people are only funny if ‘no serious injury’ is involved. So are we to assume the other 89 would only laugh if the fall resulted in a horrific facial disfigurement or similar injuries? Some involved close relatives or friends yet most were generalised as ‘people’. One went far enough to paint a picture of an ‘overly-fake tanned fat woman falling off the escalators.’ No matter how childish we may think it is, it seems watching someone fall over is a laughter inducing image.  

Although we had a large sample size (of over 3000), it was clear our key demographic was young teenagers and children, whom attended these conventions. This led to a significant percentage of ‘jokes’ being something that triggered laughter. Therefore, it is not representative of all the groups that make up the population however it can be seen as an accurate interpretation of the various triggers of laughter. 

It is safe to say that being a part of this experiment and collating this mass of data definitely ranks higher than our expectations of acting as general dogsbodies for two weeks. However we still hold the belief that we would have made excellent photocopiers.

Lucy Kent, Haafizah Khodabocus and George Ponniah                                                                                       Work experience students from Graveney School

 

INTERNET LINKS

[1] Miranda Hart http://www.youtube.com/watch?v=HFUj5GqTArU

[2] Harlem Shake http://www.youtube.com/watch?v=iGCy2zp4XuA

[3] Mr Bean Olympic Cameo http://www.youtube.com/watch?v=BSg_AvqF0qU




This is science!

posted 10 Jul 2014, 06:43 by Dana Boebinger

Check out Sophie's piece about the difficulty of studying laughter, featured on the Guardian website!


Videos from the article:

YouTube Video


YouTube Video


Videos!

posted 10 Jul 2014, 06:41 by Dana Boebinger   [ updated 10 Jul 2014, 06:43 ]

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

YouTube Video



2012 TEDxImperialCollege talk

YouTube Video



2012 UCL mini-lecture about learning to laugh

YouTube Video



2012 Sophie on James May's "Things You Need to Know - The Brain" (S02E02)

YouTube Video



2013 Sophie on "Some Boffins with Jokes" - BBC Four

YouTube Video




2013 UCL Lunch Hour Lecture - LOLZ! The science of laughter

YouTube Video


Random Acts - Reeps One

posted 7 Feb 2014, 04:38 by Sophie Scott   [ updated 7 Feb 2014, 04:40 ]

here is the Random Acts film about our work with Reeps One

YouTube Video




the behind the scenes footage is here

YouTube Video



the film was made by I Owe Youth which makes the Speech Communication Lab happy as we like puns.

behind the scenes with Reeps One

posted 21 Jan 2014, 03:57 by Sophie Scott

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!

are comedians unusual?

posted 16 Jan 2014, 09:03 by Sophie Scott   [ updated 16 Jan 2014, 09:07 ]

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.

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.

1-10 of 40