The Insight Podcast

Ep3. Hip hop, psychology, iconography

December 02, 2020 FH Media Consulting Season 1 Episode 3
The Insight Podcast
Ep3. Hip hop, psychology, iconography
Show Notes Transcript Chapter Markers

Gráinne and Louise catch up with Professor Griffith Rollefson to find out why a musicologist is getting involved in data science. Psychologist Claudette Pretorious talks about her research into how teens access mental health services online, and Professor Mathieu D'Aquin talks Insight@NUI Galway, and the role data science can play in disciplines such as music and iconography .

Gráinne Faller  0:08  
Hello, and welcome to episode three of The Insight Podcast. I'm Gráinne Faller.

Louise Holden  0:13  
And I'm Louise Holden

Gráinne Faller  0:15  
And if you can hear a strange, rumbling, farty noise in the background, that is my dog, Charlie, of Snoring Dog Productions. He really doesn't care for podcasts, but he should because we have a fascinating one lined up for you today. We're going to be talking to Griffith Rollefson, a music professor in UCC. Louise and himself got to nerd out quite a bit.

Louise Holden  0:40  
Yeah I really enjoyed hearing about Griffith work. He's looking at hip hop and its evolution around the world. I'm more on the folk side, but there's a huge amount of common ground between these two kinds of music. So looking forward to that one.

Gráinne Faller  0:52  
We also spoke to Claudette Pretorius. She is a psychologist, would you believe. We're struggling to find the data scientists in Insight at the moment.

Louise Holden  1:01  
It  turns out there's data science in everything. We spoke to Matthew D'Aquin as well in NUI Galway. He was supposed to be telling us about what he's been doing down there. But actually, he ended up talking about music, too. 

Gráinne Faller  1:10  
He did, indeed. But first, we're going to talk to Griffith Rollefson. He's a Professor of Music at UCC and an Insight funded investigator. Now if you're struggling to see how Professor of Music and data science connect, you're not alone. I was too. So I decided to ask him. 

Gráinne Faller  1:27  
Professor of Music wouldn't be necessarily the first type of person you would that would spring to mind when you're thinking of Insight SfI Research Center for Data Analytics. Is it rude to ask what exactly you're doing here?

Griffith Rollefson  1:41  
No, not at all. So I proposed this project - I don't even have I have a very minimal background in digital humanities - but James O'Sullivan, and Orla Murphy, over in the digital arts and humanities helped me put together a project. A big data project and a sort of semantic tagging, semantic web project, that looks to use, to sort of leverage those tools and connect them up with my own toolkit of sort of close listening, discourse analysis. And so I'm just looking to leverage these tools to do a more global analysis. I can do a lot of sort of fine grained analysis on, you know, a large number of tracks from my perspective, you know, in the hundreds, looking at how hip hop knowledge moves across the globe, from different languages. But I'm really interested in looking at a sort of a more massive scale, how this music has spoken to people all over the world. It's kind of a contradiction that this most local of musics would become the most global of musics. And so I use that, that contradiction as a challenge. And I think we can gain some insight using some of these techniques and sort of training AIs for example. We haven't gotten nearly to this point yet, we've just built the corpus, but training, machine learning to look for similarities in the sort of conceptual framework that I'm sort of deep into, in American, British, Irish, French and German hip hop, and then connecting up different, I've got postdocs with specialisms in African hip hop, and in East Asian hip hop and Australian hip hop, and I'll get into Brazilian and Latin American, Spanish hip hop later as well.

Louise Holden  3:33  
So it's the objective to create almost a map like semantic map. So you can you can visualize how all these different genres or sub genres have grown and connected to one another. Is that it?

Griffith Rollefson  3:44  
Yeah, exactly. So the example that I always start with is in hip hop, we have this idea of the third eye. So this is, um, hip hop consciousness. You know, it's media literacy. It's, it's wokeness. Really, I meant, but it's, it's wokeness, you know, 40 years before we had that term and hip hop. Hip hop is a very critically incisive form. And the really good stuff uses these ideas like the third eye, hip hop consciousness, but of course, the third eye is a South Indian, South Asian, excuse me, concept, the Ajna chakra of your third eye opening. And so hip hop itself is this, it uses concepts from around the world, whether it's martial arts, sort of Confucianisms are all over hip hop. Cosa Nostra, you know, like, Sicilian mob aphorisms and these kinds of things make their way into hip hop, as well as, of course, the sort of, each one teach one in different African wisdoms. And so we're not just looking at how these ideas are taken up in hip hop, and spun out to different locations again, around the world from Paris to to Berlin to to Ankara, but hip hop itself is pieced together from these bits of wisdom from around the world. There's a famous, what's the phrase, hip hop builds something out of nothing. It takes some things from everywhere, and builds them into a sort of unique thing. So that's what I'm hoping to track - is how hip hop is this site, a post colonial site, that takes knowledge from all over the world to fashion a new sort of emancipatory critical eye.

Louise Holden  5:29  
So is it largely the lyrics that you're examining? Or is it is it the production of the music as well?

Griffith Rollefson  5:34  
Yeah, that's another great question. So we're, like I say, we've just been building up the corpus, and it is, it is just lyrics for now. Bu t we've been tagging those with the ways looking at how the lyrics are often supported by sonic elements. So the beats, in fact, one of our postdocs, Jason Ng is Thai Australian and he is a dancer. He came up with hip hop breaking crews. So not just sound but also movement. So there are different dance moves that have different meaning. So right now, like I say, we're, we're building up this corpus, connecting it to sonic elements, and then eventually, we'll want to do sonic analysis, everything from beats that have particular meanings from sort of hip hop's - hip hop is deeply intertextual, you can think of sort of hypertexts websites like Rap Genius, which is now just Genius are sort of full of these, how, look at how fans deduce the meanings. And look, go down these wormholes, these these rabbit holes to find out the different meanings. And it's often deeply like clicking through links, a single lyric can open out onto a wide array of different references. So that's the that's the end goal is to look at not just the lyrics, but the sound, the images, the icons.

Louise Holden  6:57  
And tell me is it always very easy for you to identify a piece of music as hip hop? Or is it in question at times?

Griffith Rollefson  7:02  
Yeah, that's a good question. So we have had some questions about, you know, where do we set the limits? But, you know, for me, no, it's not hard. That there is there's, there's a difference between rap music, and hip hop, in some senses, and some pieces of music can be rapped, and not be hip hop. Hip hop really signifies that, that cultural element, where it's not just somebody rapping, but somebody using this form of spoken poetry over music, to dig into this archive of knowledge. So yeah, in fact, that's an excellent question because it's, um, it's not just the musical form. And ultimately, we won't be looking at just absolutely everything under the sun, but really digging down into I call it a hip hop knowledge mapping project. So whatever form those knowledges come in -lyrical, sonic, iconic, performative, that's sort of the point. to find, because hip hop has this. That's the theory. So the theory of hip hop interpolation says, these people who for whom hip hop has resonated, were already hip hop. Alright, so the idea is that hip hop is a storytelling form. But every, just like I was talking about the Ajna chakra, every just like every culture on Earth has its own legacy and cultural specificities. Every culture on Earth has had its own storytelling traditions. And it's only over the past, you know, to 300 years that literate culture has caught on. But hip hop is a sort of key back to pre literate forms of communication. The Irish bards told the stories of this nation just as the West African griots, or Chinese storytellers told their histories before the emergence of writing in a sort of, on a mass scale. So that's another one of the the ideas is saying, hip hop allows people all over the world to access those, those sort of ancient ways of knowing those pre modern ways of being, which, in many cases are much more sort of healthy and communicative in a sort of community sense, sharing community sharing knowledge

Gráinne Faller  9:21  
What brought you to Cork?

Griffith Rollefson  9:23  
The job, I was born in California, and my first job was in Berkeley at UC Berkeley for short term post and then I bounced over to Cambridge that got me over the, over the pond. And then this post came up. It was the first sort of pop culture, pop music post in Ireland. And I was offered the post, and it was really getting here to UCC and getting more in touch. The end of my end of my first book - my first book is about hip hop in Berlin, Paris, London called Flip the script, Europe an hip hop and the politics of post coloniality, But the last chapter and the conclusion of that book, I, I start looking at Cork and looking at Ireland and looking at Irish rappers. And that was when sort of the light bulb went off that you know, this this music is in Ireland especially, is so closely attuned to the I'm going to pronounce it wrong now that shinnick, Shana key, Seanchaís, the storytellers, the bards, you know, the, the harp being the the national symbol directs, right to that point about Ireland as a nation of storytellers. But the crazy thing is that is that I came up as an African American and African Americans look to the Greek to the West African storyteller. And the symbol of the Griots is also the harp, the Kora. It's a spike spike harp. And so I just thought, Oh, my gosh, here it is. The key, the Rosetta Stone.

Louise Holden  11:02  
That's thrilling. I can't wait to hear about this.

Gráinne Faller  11:06  
Yeah, we'll definitely have you back on again, and can't wait to hear more. With that said that was fantastic. Griffith, thanks so much for joining us today.

Griffith Rollefson  11:14  
Yeah, my pleasure. Thanks a lot.

Louise Holden  11:19  
Among the many activities that moved online over the lockdown that can be few more important than mental health services, young people in particular are turning to the internet to understand their symptoms and to seek help. So today, we're talking to Claude Pretorius, who's completing a PhD in computer science and inside UCD, that has a background in counseling psychology. So we're very interested to hear how how God has married these two areas under her research. Claudette, can you tell us? 

Claudette Pretorius  11:43  
Oh, yes, so I used to work as a psychologist in a university. And our university actually also experienced the lockdown, but for different reasons. And because I was in the counseling service, we obviously had to stop our services to students. And this is a few years ago. And what really interested me was how students would then go about looking to find help still, although the services weren't there. And a lot of them resorted to the internet, and reach out to us through our Facebook page, and all those and through email, and I just kind of realized there was a real opportunity like for technology to kind of facilitate services, when services couldn't be in person, for whatever reason. And so I got this wonderful opportunity to do my PhD at UCD, at the Insight Centre and through Marie Curie funding, which is amazing. And so my research just really looks at how young people use the internet and online resources to look for help for their mental health and to support their mental health. It's a fairly new area. So it was really, really interesting in that as well, that you just finding out a lot of things and finding out a lot about young people's behaviors online. And then we were also just really interested in finding out like, how we can actually design technology to make that process a lot easier and safer, because there is a lot of good stuff on the internet. But that goes back to good stuff on the internet, obviously, as with everything. And it's sometimes it's really hard to find the good stuff. And so we wanted to like kind of figure out, how do we design a technology to make it easy to find the good stuff? But also, how can people who are providing services online design the services better so that it suits young people's needs better? Yeah, so obviously very relevant for the pandemic.

Louise Holden  13:37  
I don't go off road. But can I briefly ask you why was there a locked in on the campus where you were working?

Claudette Pretorius  13:43  
We had student protests, because of fees. So, I was based in South Africa. And for a lot of students, the fees were just it's just really expensive, and they can't afford them. And so the students protested for fee reductions. And it was a nationwide protest.

Louise Holden  14:02  
Interesting. So but I suppose every cloud, it gave you an opportunity to see what happens when students are cut off from the normal way of accessing services?

Claudette Pretorius  13:42  
Absolutely. 

Louise Holden  14:11  
What has working at insight, in particular brought to your research, you're usually using data to investigate this area. How has that worked for you?

Claudette Pretorius  14:20  
So I think the biggest thing about Insight when we're talking about this area is the fact that it's so multidisciplinary. And I think when people hear the term data analytics, they think of just strict computer sciences, and AI and machine learning. And it is all of those things. But there's also just so much more, and data can just use a lot of different types of data. And so I think Insight is this great case, or this melting pot of different disciplines and different people who you wouldn't normally get to work with if you were kind of stuck in the silo of your discipline. So traditionally, like psychologists work with other psychologists and have a very psychological way of thinking about everything. But when you're kind of working with computer scientists and human computer interaction people, and even like physios, like it just changes your way of thinking. And it exposes you to different ways of doing research different ways of gathering data even, and analyzing data, because you were trained to in your own discipline. So I think that's the great thing about, about Insight. Yeah,

Louise Holden  15:28  
If you can, can you give us a sense of what your research might look like for a student who ultimately benefits from it, if somebody decides to sit down in front of a laptop and seek mental health support,

Claudette Pretorius  15:39  
My research will kind of pan out in two practical ways. One way is we'll develop a tool. So it's kind of like a search tool that'll help like a young person or a student find the mental health resource they're looking for, based on the type of content they're looking for, whether they're looking for like videos, or blogs, or whether they're looking to chat to someone online or offline. So we're looking at designing that kind of technology. And then on the other side, we're making guidelines available to mental health providers. So people like Jigsaw and Spun Out who have an online presence, and just kind of like, give them guidance on how to best design their online sites, and to let them know, look, this is what young people are looking for, if you're wanting to meet their needs, try and design for these kinds of things. So we're kind of hoping that if you marry those two together, that their online experience is a lot more satisfactory.

Gráinne Faller  16:38  
What insights are you getting at the moment to to how young people seek help online?

Claudette Pretorius  16:43  
Yeah, so young people are obviously the first port of call for lots of people Google, just because Google's kind of like we all do it. If we have a question, we'll just pop a question to Google. So lots of young people do use Google. The problem with Google is, if you don't use the right keywords, the right results don't necessarily come up. Other young people try and access things through their college or school. So they'll go on to their college website and see what kind of resources the college recommends. And then the other thing is, yeah, lots of people kind of know about people like Jigsaw, Spun Out or Pieta House. And they'll go directly to those places, and then go and find resources through that. But I think the interesting thing about a Google search is that when they do that, they open like 10 or 20 tabs, and they'll go through all of them quickly. And they'll find the most relevant one and close the ones that don't look relevant. And then what they also do is they cross check that information. So like if one side is saying this, they'll check that against something else. So it also kind of shows that there's, they're discerning about the information they are reading online, which I think is really important as well, so, 

Louise Holden  17:53  
And do you differentiate between people simply doing research into what they think their mental health issue might be on people who are actually seeking a service online? 

Claudette Pretorius  18:03  
Yeah, so we found in one of the studies we did, we found that there's probably three types of help seekers, there's the person who's going online purely just for information, and they don't want any sort of outside interference, they just want to go through the information, learn more kind of figure things out for themselves. And then there's the second type who's the more person centered help seeker, and they're looking to actually connect with someone, whether that be through a personal story, or through chat, or, or an email, like connecting with another person, and they're kind of looking for more direct, sort of interpersonal help. So you're first helps the kind of the person who just wants to help themselves, they don't, you know, that's fine. They're very self reliant. But the second type of person actually wants help from another person. And then the third type of person we found is the crisis help seeker. So that's someone who's in severe distress, he's having a really tough time and just wants to connect with someone to have a chat, and just help them through that moment.

Gráinne Faller  19:04  
When you come to the end of this research, Claudette, what do you see as being next for you? would you would you go back to being a counseling psychologist? Or do you think you might continue your studies? Or is that? Is it all way too early to be even thinking about?

Unknown Speaker  19:17  
No, it's not. So I'm actually starting a postdoc with Insight. So I'm going to stay on at Insight, which is great. I'm really excited about that. And I think the next thing is to build something so that the whole my whole PhD was kind of exploratory and finding out what it is young people are doing and what they want. And we did build a very low fidelity prototype as part of the last study that I did, which we tested, which revealed some really interesting findings, but I think my future research is going to like get is going to focus on maybe building something a bit more concrete a bit more sophisticated and seeing how that works. And just testing that out. And just looking at more collaborations with local organizations because it's no use and buildings. thing, and no one can use it. So also just seeing how we can kind of partner with local youth mental health organizations and just see if that could be something that was worthwhile to them

Louise Holden  20:11  
Can I ask you? How much have you had to read yourself into the technology side of the computer science? And have you enjoyed that aspect?

Claudette Pretorius  20:18  
So yes, I cannot program anything. And I think that's fine. I think that there are people who have much better skills than I do at those kinds of things. And I think that's why collaboration is so important. And that's why working in a place like Insight is really, really important because you can find partners for collaboration so easily. And, and I think that's important. I have learned a lot of skills from the human computer interaction domain, which I think is great. And which I've really enjoyed, because it's a different way of thinking it's a very different way of doing things. But the more hardcore computer science skills, I think are above my paygrade.

Gráinne Faller  20:59  
But it's been fascinating talking to you, thanks so much for joining us today. 

Gráinne Faller  21:06  
And finally, we spoke to Matthew D'Aquin director of Insight at NUI Galway about his plans for the site.

Louise Holden  21:13  
So just, I mean, obviously, things are going to be a bit unsettled for a while, yet, but Insight 2 is in full flow and projects were continuing and starting underneath that new banner. What are you looking forward to under Insight 2? And what do you see what kind of projects you see maturing over the next 12 months or so?

Mathieu D'Aquin  21:31  
Well there are a lot. I tend to have, in terms of research myself a bit more focus on all the aspects related to data engineering, which have been growing enormously in Insight in the last few months with with a push and with BD, industry partners and public sector partners, realizing that those aspects are very important. So within insights, we have been more and more talking to a large variety of people, including the people working in care of crucial analytics, in theory of climate change, we have, you know, the new platform research initiatives, which have been started with which we are very much involved. And a number of them have been growing with this aspect of data engineering fairly much at the center of let's say, at the start. The data engineering aspects are ones I care a lot about, because they are often described as the heavy lifting, so things without which you can't go into doing the clever AI and the clever machine learning. So there are really, really very good use cases coming from a large variety of domains that are, say growing within Insight currently, and there will be a lot of very interesting results coming out of it.

Gráinne Faller  22:43  
Do you have any examples? 

Mathieu D'Aquin  22:44  
Yes, the examples are very varied so that's why it's very hard for me to pick one. So there is one I liked a lot and I care a lot about is we we just got a new project accepted the European project, which is about mining musical data. So it's with partners from the Netherlands, UK, and Italy and France, and we are going to do is look at, say a large, large collections of music data, which is either data about music or actual music data - so things like MIDI and sound - in order to be able to extract patterns that will help understand the evolution of music, understand how different pieces of music relate to each other, and how music say has evolved in time through those specific collections. And one reason why am I specifically excited about this is I started getting into this based on an initial collaboration with the Irish Traditional Music Archive some time ago. And and we are going to work with the Irish music archive to especially look at patterns of music within traditional Irish music. And again, this is an example of how, say the often described as terribly boring and technical skills we have within Insight can end up having an impact on on things which of which are much more exciting and much more important, for sure.

Louise Holden  24:18  
Yeah, and I will have to explore that again in greater detail in a future podcast because I know there'll be lots of people across inside who'd like to hear more about that we I think we've a fair few musicians among our ranks.

Mathieu D'Aquin  24:28  
Yes that's correct, then it's interesting that you know, for a French person like me to see how music is so important in Irish peoples' lives and how a place which is profoundly made of - I'm not using that term in a negative way - made of nerds, is also a place which is full of musicians and artists.  It's really this connection between the two which is very exciting in the sense that it's, it feels unusual, and but when you look into it, it's not that much. We actually have a small local initiative in in Galway around looking at this connection between arts and music. We started before the lock down with the idea that what we wanted to do is somehow find ways to get exhibits that will show in an artistic ways are kind of things we do. And one of the projects we're exploring is, also one on music and on, basically, sonifying music. And say, I'm absolutely amazed by how, when you start talking about this, with computer scientists, they surprisingly get excited. But also, when you start talking about these musicians, they also get excited. The connection between the two gets, gets something which is really, really quite amazing, frankly.

Gráinne Faller  25:50  
It is interesting, the fact that I mean, over the years, as expertise and research has become more refined things have tended to get a little bit siloed. And we're beginning to see that re integration of disciplines. And it's it's fascinating, actually, because even looking at the list of the new FIs in Insight 2 I mean, there are so many, I mean, there's music professors there, there's English professors, they're just people who are in disciplines, you wouldn't expect them to find in, you wouldn't expect find them in Insight, I guess?

Mathieu D'Aquin  26:18  
No, that's correct. And I say, it's, it's actually interesting, because if I look at the common point, between all the many projects I have had in the last few years, they have no common point in terms of domains, mostly, because the one thing they have in common is is getting with, to kind of, to work with somebody who does not work on computer science related aspects in the past. I always tell the story, and it's actually an interesting story, because it's, it's one that is finally getting into a conclusion. Many, many years ago, I was approached by a professor in iconography, actually, she was a lecturer at the time, who had a project on classifying and recording representations of hell, in churches in Crete from the 13th century. And you think, why would I want to do that? Right? I mean, I'll say that it felt so very specific, so very far away from what we usually do. And the first meeting I had with her, the first words she ever told me, was, 'I live in the 15th century, and I'm very happy about that,' she didn't care or want to know about technology. But I think I have been working with her for six years now. We are still in regular contact. And she just asked her publisher, to send me the two volumes she has written with, with our colleagues and with the team, on that particular project about the representations of hell in Cretan churches from the 13th century. And the funny thing is, I know a lot now about how you represent a thief, or any kind of sinner, and what kind of punishment they might get. But also I learned a lot about how technologies apply. And she learned a lot about, you know, the 15th century being great, but there are some stuff from the 21st century that can help as well. So it's, it's, you know, it's the kind of story that makes things really what makes, you know, again, our often seen as boring job, get a bit more exciting, and advance our our research with some effort, but but it generally pays off.

Louise Holden  28:40  
Well, there's something inherently poetic, about reaching across time like that using using technology from right now to lift the lid on the lives of people from 400 years ago. And those kinds of synergies are happening across Insight. I mean we see it with the work that Derek Greene and Susan Levy are doing in UCD, with the British Library, and you know, the history of vaccinations and these kinds of issues. And once you start to see that there's a magic that comes in the space between not just the chronological space, but also the different disciplines coming together and people learning from one another. It's wonderful.

Mathieu D'Aquin  29:17  
It is quite amazing. As long as it's done with a certain level of humility. I also remember the first time it ever happened to be and my contribution was - and the first time it didn't work - because by the time I realized that as many other computer scientists, I tend to be the guy with a hammer looking for a nail, and not really sufficiently listening to what is the actual problem they have. Instead of trying to fit my solution to, to their project. So it does it does require effort. But yeah, it's quite amazing when it works.

Gráinne Faller  29:52  
That's a really interesting point, actually, so much of the discussion about Insight 2 has been about how to encourage researchers to collaborate with one another, how are we going to find those synergies across the organization and that, and what you've said there about listening is such an important thing. You know, when you're an expert in your field, and you have so much knowledge, maybe the instinct isn't necessarily to sit down and really listen to the person across the table from you.

Mathieu D'Aquin  30:18  
It's entirely true. And it's actually somehow even more true between disciplines of data science or computer science. It's very hard for somebody like me, you can come from a tradition of symbolic AI, with, you know, explicit reasoning rules and, and, you know, knowledge representation, to talk to somebody who has been doing neural networks their entire life, because we just don't speak the same language, we don't solve the same problems. And our frameworks just don't align, they don't work with each other. So there is, this added difficulty, that not only we don't, we tend to be as far away from each other, as I would be from from somebody studying iconography. But also, the complementarity is not necessarily as clear because both disciplines, say implicitly tend to have a bit of competition between them. It's it's, you know, it's a question of sometimes, and that's the wrong way of looking at it,  it can be a question of who is right and which one is better. And, and that really doesn't help collaboration. So but there are obviously ways to deal with this. But, but it does take a long time. My previous boss always says it's much harder to work with somebody who is very far away, than with the guy just next door.

Gráinne Faller  31:43  
Very interesting point. Isn't thank you so much for talking to us today. That was fascinating. Thanks, Matthew. And that is all we've got time for today.

Louise Holden  31:56  
If you enjoyed the show, please do rate and review and subscribe. If you'd like to get in touch with us. You can get us at info@fhmediaconsulting.com. We'd love to have you on the show if you've got something interesting to talk about. And please join us next week for another serving of great science straight to your ears every Wednesday.

Louise Holden  32:19  
This has been a snoring dog production on behalf of the Insight SFT Research Center for Data Analytics

Transcribed by https://otter.ai

Griffith Rollefson
Claudette Pretorius
Mathieu D'Aquin