Approaching Feminism as a Male Data Scientist

Blog post by WiMIR mentor Ryan Groves, co-founder of Melodrive

11-12-abstract-painting-of-different-trees-reflected-in-watercrop

Painting Equality used with permission from artist Osnat Tzadok, find original here.

The internet has provided a new platform for an obscene amount of information. Anyone with a computer and a connection can now be heard in the international community. Through the accessibility of information, citizens have become journalists, comedians, celebrities, laughing stocks, community leaders and even scientists just through the means of access to this information tsunami. One particular aspect of this is the ability for marginalized groups to directly confront those who are more privileged, or even perpetrate that marginalization.

I am a man. I had a relatively privileged upbringing–good schools and active parents. When I first started reading about feminism, I had the knee-jerk reaction that most men have: “They couldn’t be talking about me”. Surely I was a more-informed member of the opposite sex. Surely all of my accomplishments were based on my merit, and not on some advantage bestowed unto me. Surely I wasn’t gaining those advantages through the disadvantages of my peers.

As a data scientist, I teach machines how to understand human problems, and to perform human-like solutions. I decided to do what a good data scientist would do–to gather evidence, form a hypothesis, train the model to solve that problem, and test. Only this time, I’d be the subject.

Gathering Evidence

Let’s talk about the elephant in the room: Mansplaining. According to the New Republic [1], the essay that kicked off the whole idea was by the decorated author, Rebecca Solnit, in her piece, “Men Explain Things to Me; Facts Didn’t Get in Their Way”. In it, she says that “the out-and-out confrontational confidence of the totally ignorant is, in my experience, gendered. Men explain things to me, and other women, whether or not they know what they’re talking about. Some men”. Even in my marginally-more-enlightened state, after having opened myself up to this issue, I still find my instinctual male reaction kicking in when I read this sentence, which is to reject this notion outright. But that is exactly the point–why would I not approach this as evidence, as opposed to rejecting it as opinion? The author continues:

Every woman knows what I’m talking about. It’s the presumption that makes it hard, at times, for any woman in any field; that keeps women from speaking up and from being heard when they dare; that crushes young women into silence by indicating, the way harassment on the street does, that this is not their world. It trains us in self-doubt and self-limitation just as it exercises men’s unsupported overconfidence.

As a man, this was initially a foreign idea to me–an issue of which I was totally unaware.

The problem goes beyond Mansplaining. More recently, I read a really insightful blog post by a female engineer & designer, Nikki Lee, titled “Ride Like a Girl”. In it, she relates the experience of living as a woman in a world designed for men to being a biker in a world designed for cars. In it, she says:

Maybe you’ve noticed that cars are kind of scary. They are more or less scary depending on where you live — if you’re on a nice calm street in Seattle, drivers will give you plenty of room and wait patiently for you. Other places, drivers will try to run you off the road. And no matter where you are, you know that the cars around you could really mess you up if something went wrong.

[…]

In places with great cycling infrastructure, it’s really easy to get around just as efficiently as driving. But most cities don’t have that. These environments aren’t just not built for you, they’re constructed in a way that actively excludes you.

This analogy really resonated with me, because, like most people, I’ve had this experience while biking. I have indeed witnessed negligent drivers who impose the risk of real physical harm on bikers through their negligence. I have been held to a double-standard by having those same negligent drivers call me out for minor traffic infractions.

In industry, there are plenty of examples of the same disadvantages. I recently attended a Meetup event for gaming and game development in Berlin, and had the honor to hear a talk given by Brie Code. Brie Code is a speaker, writer, and the CEO of a new game studio, Tru Luv Media. She has worked for years in the games industry as a game AI developer and designer. Brie had always found that she had a very different response to games than traditional gamers. As she presents in her blog post on the topic:

[M]ultitudes of white masculine gamers who dominate the games industry have made experiences that are relevant to them but not to most people.

Brie often didn’t feel the “fight-or-flight” response to dangerous in-game situations. When she tried to explain this to other game developers and executives, they would confront her about it, even indicating that her own experience was wrong. Brie decided to do some research into the underlying assumptions that fuel those opinions. The research she discovered was compelling.

The most-commonly-known physiological response to threatening situations in games is fight-or-flight. As Brie Code puts it:

With fight-or-flight, your sympathetic nervous system kicks in and releases adrenaline followed by dopamine. If you like games like this, it’s probably because adrenaline and dopamine are very enjoyable. Your pupils dilate. Your heart beats faster. Your airways open up. And you feel exhilarated. You feel alive. You feel powerful.

Because Brie (as well as many of her friends) did not respond to games in this way, she dug up the psychological research around these responses, and found a completely different, competing response: tend-and-befriend. As Brie describes:

If you experience tend-and-befriend, it’s because your body releases oxytocin or vasopressin when you’re stressed, followed by opioids. This calms your sympathetic nervous system so you don’t get the flood of adrenaline. Instead of wanting to fight or to flee, you stay relatively calm, but aware. Your pupils dilate, you become fearless, and you are less sensitive to pain. You instinctively want to protect your loved ones, to seek out your allies, and to form new alliances. Oxytocin intensifies social feelings, and opioids feel extremely warm and lovely.

Research shows that oxytocin is involved in the physiological response to these threatening situations more-so for women than for men. In order to discover why this tend-and-befriend response was so overlooked, Brie did even more research. Prior to 1995, only 17% of the participants of psychological stress studies were female. In addition, according to an article by Think Progress, a 2014 review of sports and exercise studies showed that only 39% of 6 million participants were women. Brie found that this bias existed in things such as air-bag deployment, or identifying the signs of heart attacks.

However, it’s not only women that respond to games with tend-and-befriend. It seems that there is simply more to the picture than the traditional fight-or-flight response, which is exactly what Brie set out to prove.

Even in the highest office of the (American) land, female staffers were feeling overrun by their male counterparts. So much so, that they devised a strategy of mutual affirmation.

If we’re forming our opinions based on biased information, and excluding opposing opinions, then even what we consider to be fact is potentially false. It certainly seemed there was a common theme in all of these examples, which was succinctly described in the previously-mentioned article by Nikki Lee: These environments aren’t just not built for you, they’re constructed in a way that actively excludes you.

Forming a Hypothesis

I decided to challenge myself. Perhaps, rather than hypothesizing that all of this evidence was untrue, and trying to prove that, I should do exactly the opposite: I chose to hypothesize that all of this was true, and challenged myself to prove that it was.

Training myself to be aware

Since making this decision I’ve tried to be more aware of the issues facing women specifically. But how does one really change her own perspective? It isn’t solvable with a single bout of effort. It needs to be as consistent of an effort as it is for those who are not at a natural advantage in society’s constructed environments–one needs to be as aware as a biker in a city. As has been evident with the recent political happenings in the USA, sources of information can have a huge effect on your thoughts, habits, and perspectives. The two camps for Clinton and Trump were both guilty of consuming information that supported their own perspectives, which was why it was such a rude awakening for the Clinton supporters when Trump was elected. In general, we think of receiving information as a passive process, and in many cases it is hard to avoid consuming certain information. However, there are ways to customize your acquisition of information, and to control the perspectives that you choose to expose yourself to. In effect, you can run your own self-inflicted propaganda campaign.

I also endeavored to prove the hypothesis by gathering my own data. I found there to be ample evidence of the gender-based slights in my daily life. A female coworker filed a sexual harassment complaint, and the whole company was subjected to a sexual harassment seminar–with the glaring exclusion of the offender. Another female colleague was asked her opinion on her area of expertise, only to have them initially reject her opinion outright, and then claim her idea as their own in the final version. In a pickup basketball game, a female teammate suggests reducing the game to 4 on 4, because we only have teams of 5 and 6, and it would be hard for her team (of 5) to be running for the entire game with no substitutes. A man responds, insisting it be 5 on 5, and shouts “we’re men, and we want to run!”.

I’ve even noticed simple micro-aggressions in the academic world. At a recent conference, I walked up to a particular poster presentation in the middle of a conversation between the presenter and a female participant. Immediately, the presenter’s focus shifted from the woman to me, and it was as if the woman was no longer part of the conversation.

Testing

With the mounting evidence from both personal experience and outside sources, it seems that gender issues are real, and pose a serious barrier to equal opportunity. So how do I use this gained perspective to make a real change?

Spread the Word

One constructive approach is to evangelize my own personal process, and to encourage other men to change their perspective. It is indeed both genders’ responsibility that people are treated fairly. This blog post is my first attempt at that. But it also involves ongoing discussions and dialogue.

Confrontation

Now that I am convinced of my hypothesis, it’s my responsibility to confront micro-aggressions and unfair treatment that occur in my surroundings.

Continue My Personal Awareness Campaign

My perspective is not perfect, and I don’t expect it ever to be. For the last 2 years, I’ve been a mentor for the Women in MIR (WiMIR) mentoring program. It’s a fantastic initiative that is trying to address some of these gender-based discrepancies in academia, with a specific focus on the MIR (Music Information Retrieval) community. Part of the reason that I like to be a WiMIR mentor is to simply be aware of the issues that women face in the workplace from their perspective, since there are always things that I personally overlook coming from the perspective of the opposite gender.

Build a Positive Work Environment

As a business owner in my capacity of co-founder of Melodrive, I have the unique opportunity to sculpt my own business’ work environment, and it’s my responsibility to create an environment that is as inclusive as possible. Recently, the Augmented Reality company Magic Leap had a lawsuit alleging gender-based discrimination. In the lawsuit, there contains a really thoughtful list of recommendations for a company that wants to be more inclusive (written by Tannen Campbell).

  • Creating “parents’ guides” for cities where Magic Leap had offices, and listing recommended doctors, babysitters, etc. for newcomers
  • Establishing internships for female students to code/build for mixed reality
  • Creating an internal mentorship program pairing men with women;
  • Increasing paternity leave to show that Magic Leap values a man’s role in child caregiving and allows busy moms to return to work sooner if they choose;
  • Tying diversity to extra bonuses, e.g., giving a department head an additional bonus for increasing gender diversity
  • Implementing mandatory “unconscious bias training” for leadership;
  • Demonstrating a commitment to encouraging young girls’ interest in tech education by replacing the “swag”–T-shirts, notebooks and the like–handed out to important visitors with a donation in the important visitor’s name to Girls Who Code, and organization devoted to overcoming the gender gap in tech jobs.
  • Creating and appending an “Equal Pay Guarantee” stamp or seal on all Magic Leap’s position descriptions in recruiting material and elsewhere, evidencing Magic Leap’s commitment to equal pay for equal work within the same tier
  • “Fishing where the fish are” by recruiting employees from a list of colleges, provided by Campbell, with the highest number of women in STEM (Science, Technology, Engineering and Mathematics) departments

With all of these new data points, and my somewhat-improved perspective, it will hopefully be easier to be proactive to create a non-hostile work environment and accommodating employee policies as a small business owner. Additionally, perhaps I’ll be able to move the needle to a society that is more open and inclusive. And while I feel that I have personally made progress in being more in tune with the perspectives of others, I don’t feel that I’ve completed my task. This is an ongoing challenge, just like it is for those who experience disadvantages in their daily lives and beyond.

  1. https://newrepublic.com/article/118555/rebecca-solnits-men-explain-things-me-scourge-mansplaining

Ryan Groves is a computational music theorist, composer and data scientist. He received his B.S. in Computer Science from UCLA, and continued on to complete a Master’s in Music Technology from McGill University. As the former Director of R&D for Zya, he developed a musical messenger app that automatically sings your texts, called Ditty. Ditty won the Best Music App of 2015 from the Appy Awards. In 2016, his research in computational music theory was awarded the Best Paper at the prominent music technology conference, ISMIR. With his new venture, Melodrive, he and his co-founding team of two PhDs in Music and AI are looking to build the world’s best artificially intelligent composer, and to change the way music is experienced in video games and virtual environments.

“Finding courage in encouragement” by Amy V Beeston (WiMIR mentee)

Post by Amy V Beeston, one of the mentees of our mentoring program:

I am happy to have been invited by Women in MIR to blog about a recent paper, co-authored with Lucy Cheesman and Liz Dobson, which I presented at the last Digital Music Research Network (DMRN) meeting in London. Our paper reported on a community-building project underway in the north of England, which aims to support and encourage women and girls in (and into) academic and industrial careers relating to music technology.

We were recently awarded a Catalyst: Festival of Creativity grant as part of a city-wide Year of Making celebration in Sheffield. This allowed us to run a series of 8 expert-led workshops, peer-learning maker-space days and a monthly social gathering. These provided ongoing opportunities for women and girls who may have felt excluded or uncomfortable in male-dominated environments to meet, share knowledge formally and informally, and thereby develop their technical and creative skills. Averaging 10 registrations per workshop, our sessions covered a wide range of topics: sound synthesis, machine listening, performance hardware, electronic prototyping, live coding, data sonification, looping and DJing. Despite being labelled as beginners’ workshops, over two-thirds of participants reported some pre-existing familiarity with the topic through their work or studies. Though participants’ lack of confidence was further apparent in their self-rated knowledge of the topic on arrival to the workshop, by the end of the session, these scores had improved in every case (Beeston et al., 2016).

It is clear from the statistics that more work to encourage women and girls is needed in the UK as well as elsewhere. UK national and international publication patterns, for instance, look very similar. We found that female-led contributions to DMRN meetings averaged around 12% of all contributions in the period 2011 to 2015 (Beeston et al., 2016), comparable to the 14% level reported internationally at recent ISMIR conferences (Hu et al., 2016). Indeed, since so few girls have been applying to study music technology in recent years (Born and Devine, 2015) or have been following careers in related industries (see e.g. female:pressure’s recent reports), a lot of energy is being spent now, worldwide, on discussions of gender and diversity in our field.

Being one of many women with an interest in both music and technology, I am of course very happy to add my own energy to this movement. But on a personal level, I felt somewhat underqualified to tell this story: I have worked with sound all my life, not with questions of gender and diversity! Nonetheless, I felt it was important to present our work at DMRN – speaking to a room comprised mostly of men – since we otherwise risk the issue being viewed as a ‘women’s problem’ rather than a problem to be tackled by our research community as a whole.

I hope therefore that our DMRN presentation will provide a useful starting point for a fruitful discussion about what the UK research community as a whole can do to help provide greater support for women and girls both entering and staying in this field. Furthermore, I am pleased to share our story with the international community too via this WiMIR blogpost, and hope that the community-building approach we have followed with the Yorkshire Sound Women Network may provide a useful model for others to adopt in order to increase the participation of women and girls in sound and music technology in other localities.

REFERENCES

Born, G. and Devine, K. (2015). Music technology, gender, and class: Digitization, educational and social change in Britain. Twentieth-Century Music, 12(02), 135–172.

Hu, X., Choi, K., Lee, J. H., Laplante, A., Hao, Y., Cunningham, S. J., & Downie, J. S. (2016). WiMIR – An informetric study on women authors in ISMIR. In Proc. 17th International Society for Music Information Retrieval Conference, 765–771.

Beeston, A.V., Cheesman, L. and Dobson, E. D. (2016). Community-building to support and encourage women and girls in music technology. Digital Music Research Network One-day Workshop (DMRN+11), London, 20 Dec.