Social media and science communication

The following article was originally intended for an Indian publication but I withdrew from the commission because I couldn’t rework the piece according to changes they required, mostly for lack of focus. I thank Karnika Kohli and Shruti Muralidhar for their inputs.

Since the mid-20th century, the news-publishing industry has wielded the most influence on people’s perception of what science is, what its responsibilities and goals are, and what scientists do. The internet changed this by disrupting how news-publishers made money.

In 2012, The Hindu used to sell a copy of its newspaper in Chennai for Rs 4.50 (or so) while it used to cost the publisher Rs 24 to print each copy. The publisher would make up the deficit by soliciting and printing ads from advertisers in different parts of the newspaper. The first major change in this regard was Google and the new centrality of its search engine to exploring the internet. Sites were keen to have their pages ‘rank’ better on search results and began to modify their content according to what Google wanted, giving rise to the industry of search-engine optimisation.

Second, Google AdSense allowed websites to run ads as well as advertisers to target specific users in line with which websites they visited and their content consumption patterns. Third, once Google News started becoming a major news aggregator, news sites re-tailored their content according to its specific needs, including reinterpreting the news in terms of the preferences of Google News and its users.

Fourth, bandwidth became cheaper around the world but especially in India, reducing the cost of accessing the internet and bringing more people online. In response, social media platforms — especially Facebook — began to set up walled gardens to keep these users from leaving the platform and consuming the news elsewhere. And when traffic to sites plummeted, their ads-based revenue came crashing down.

The effect of these ‘gardens’ has become so pronounced that recently, a paper in the journal Experimental Economics found that college students who went off Facebook consumed less news. This conclusion suffixes the belief that most people, especially in the 18-24 age group, consume the news on social media platforms with the notion that they don’t consume news anywhere else.

In another instance, Google at long last become a walled garden proper in August 2019: the fraction of its users who consumed the news on the site itself instead of following a link through to the publisher’s site had breached the 50% mark.

Finally, because the social media made it so easy to share information, citizen-journalism became more appealing, even lucrative. At the same time, social media platforms, which constantly evolve to accommodate their users’ aspirations, began to chip away at the need for public-spirited journalism. As a result, the amount of ‘bad information’ in the public domain exploded even as people become more unwilling to acknowledge that this was all the more reason society needed good journalists.

Obviously all of this is bound to have profound implications for how social media users perceive science. But while this isn’t easy to gauge without a dedicated, long-term study, it is possible to extrapolate based on what we know from anecdotal experiences. Through this exercise, let’s also move beyond the logistics of using the social media well and spotlight the virtues of getting on these platforms that so many people love to hate.

Broadly, social media allows users to organise information in a fixed number of ways but doesn’t give users control over how they are displayed. This limitation is good because the platform sidesteps the paradox of choice and forces users to focus only on what they are saying. But it is also bad because the limitation eliminates diversity of presentation, sometimes forcing users to shoehorn an idea into a note or image when a longer article or an interactive graphic would work better.

Second, social media platforms incentivise some user behaviours over others, which then constrains how users can present scientific results.

These two arcs are united by the fact that these platforms have socialised the consumption of news (and the production as well to some extent). That is, users discover a lot of news these days in social settings, such as in conversation with other users or in the timelines of accounts they follow. Such discovery happens after the news has been filtered through the lenses of others’ interests, encouraging users to follow users whose tastes they like and views they endorse, and stay away from others. This tendency is psychologically rewarding because it contributes to building the echo chamber, which is then economically rewarding for the platform’s owners.

All together, the social media — comprising platforms whose motive is profit and not social and psychological wellbeing — are populist by design. They privilege popularity over accuracy and logical value. In this regard, it would be hubristic to assume that the public perception of science has been separately or distinctly affected by general social-media use patterns.

Then again, these patterns have also helped mature the old idea that public debates aren’t won or lost on the back of strong scientific evidence or clever logical arguments. More generally, science communication in India is becoming more popular at the same time Indians are becoming more aware of the socio-political consequences of our digital lives and worlds. This simultaneity has the potential to birthe a generation of more conscientious and social-media-savvy science communicators that can devise clever ways to work around apparent barriers.

For example, scientists can adapt an app that has been designed to communicate speed, say by allowing users to rapidly compose and share text, pictures or videos, to meaningfully convey changes in that speed. They could highlight how different parts of a long experiment can proceed at different paces: sluggishly when growing a bacterial culture overnight and rapidly when some chemical reactions with it produce results in seconds.

Communicators can also ‘hack’ social-media echo chambers by setting up small, homogenous online communities. According to one 2018 study, such groups can “maximise the amount of information available to an individual” according to their preferences. The study argues that such “homophilic segregation can be efficient and even Pareto-optimal for society”.

Finally, the limits on how users can organise and present information has in fact incentivised those who had stayed away from communicating science for lack of time and/or resources to sign up. Maintaining a blog or writing articles for newspapers can be laborious. Additionally, writing for the press — the historically most common way to communicate scientific knowledge outside of journals — also means using at least a few hundred words to set readers up before the author can introduce her idea.

But if you discover that a paper has made a mistake or that you want to explain how something works, you post a few threaded tweets on Twitter in a matter of minutes and you are done. A Facebook note wouldn’t take much longer. Instagram even gives you the added benefit of using a large visual prompt to grab users’ attention. WhatsApp introduced the power to do all of this from your smartphone.

One remarkable subset of this group is traditionally underprivileged science workers (to use a broader term that encompasses scientists, postdoctoral scholars and lab assistants). While journalists are typically expected to be objective in their assessment, they — like almost everyone else — have been fattened on a diet of upper-caste men as scientists. So in the course of shortening the distance between a communicator and her audience, social media platforms empower less privileged groups otherwise trapped in a vicious media cycle, which renders them more obscure, to become visible.

Of course, some platforms exact a steep psychological price from users of currently or formerly marginalised groups (including women, transgender people, transsexual people, and pretty much everyone that doesn’t conform to heteronormativity) by forcing them to put up with trolls. So their continued presence on these platforms depends on the support of their institutions, other scientists and science communicators. And should they persist, the rewards range from opportunities to change users’ impression of who/what a scientist is to presenting themselves as a more socially just set of role models to aspiring scientists.

Obviously populism has downsides that are inimical to how science works and how it needs to be communicated, such as by falsely conflating brevity with conciseness and objectivity with neutrality. But it is always better to have a bunch of people using the social media to communicate science while being aware of its (arguably marginal) pitfalls than to have them avoid communicating altogether. This also seems to be the prevailing spirit among those scientists who recognise the importance of reaching the people, so to speak.

Science communication is becoming increasingly popular as an interdisciplinary field of its own right, wherein scientists and sociologists team up to determine the general principles of good communication by examining why some stories work so well among certain audiences, how psychological and linguistic techniques could play a part in establishing authority, etc.

These efforts parallel many scientists taking to Facebook and Twitter, posting updates regularly including comments on the news of the day (at least from their points of view) and offering non-scientists a glimpse of what it is like to be a working scientist in India. Easier access to their views also allows science journalists to contact scientists to understand which developments are worth covering and to solicit comments on the merits of a study or an idea.

In effect, Snehal Kadam and Karishma Kaushik wrote in IndiaBioscience, “social media discussions and opinions are playing a key role in Indian science. This is evident on multiple fronts, from increasing accessibility to administrators and enforcing policy changes to determining the way the Indian science community wants to be represented and viewed, and even breaking down silos between scientists and citizens.”

There are many resources to help scientists understand the social media and use these platforms to their advantage — whether to popularise science, find other scientists to collaborate with or debate science-related issues. I don’t want to repeat their salient suggestions (but @IndScicomm is a good place to start), plus I am not a scientist and I will let scientists decide what works for them.

That said, it is useful to remember that the social media are here to stay. As Efraín Rivera-Serrano, a cell-biology and virology researcher, wrote on PLOS, “These platforms are shaping the future of science and it is imperative for us to exploit these avenues as outreach tools to introduce, showcase, and defend science to the world.”

A view of a road from next to a gutter filled with leaves.

Two sides of the road and the gutter next to it

I have a mid-October deadline for an essay so obviously when I started reading up on the topic this morning, I ended up on a different part of the web – where I found this: a piece by a journalist talking about the problems with displaying one’s biases. Its headline:

It’s a straightforward statement until you start thinking about what bias is, and according to whom. On 99% of occasions when a speaker uses the word, she means it as a deviation from the view from nowhere. But the view from nowhere seldom exists. It’s almost always a view from somewhere even if many of us don’t care to acknowledge that, especially in stories where people are involved.

It’s very easy to say Richard Feynman and Kary Mullis deserved to win their Nobel Prizes in 1965 and 1993, resp., and stake your claim to being objective, but the natural universe is little like the anthropological one. For example, it’s nearly impossible to separate your opinion of Feynman’s or Mullis’s greatness from your opinions about how they treated women, which leads to the question whether the prizes Feynman and Mullis won might have been awarded to others – perhaps to women who would’ve stayed in science if not for these men and made the discoveries they did.

One way or another, we are all biased. Those of us who are journalists writing articles involving people and their peopleness are required to be aware of these biases and eliminate them according to the requirements of each story. Only those of us who are monks are getting rid of biases entirely (if at all).

It’s important to note here that the Poynter article makes a simpler mistake. It narrates the story of two reporters: one, Omar Kelly, doubted an alleged rape victim’s story because the woman in question had reported the incident many months after it happened; the other, the author herself, didn’t express such biases publicly, allowing her to be approached by another victim (from a different incident) to have her allegations brought to a wider audience.

Do you see the problem here? Doubting the victim or blaming the victim for what happened to her in the event of a sexual crime is not bias. It’s stupid and insensitive. Poynter’s headline should’ve been “Reporters who are stupid and insensitive fail their sources – and their profession”. The author of the piece further writes about Kelly:

He took sides. He acted like a fan, not a journalist. He attacked the victim instead of seeking out the facts as a journalist should do.

Doubting the victim is not a side; if it is, then seeking the facts would be a form of bias. It’s like saying a road has two sides: the road itself and the gutter next to it. Elevating unreason and treating it at par with reasonable positions on a common issue is what has brought large chunks of our entire industry to its current moment – when, for example, the New York Times looks at Trump and sees just another American president or when Swarajya looks at Surjit Bhalla and sees just another economist.

Indeed, many people have demonised the idea of a bias by synonymising it with untenable positions better described (courteously) as ignorant. So when the moment comes for us to admit our biases, we become wary, maybe even feel ashamed, when in fact they are simply preferences that we engender as we go about our lives.

Ultimately, if the expectation is that bias – as in its opposition to objectivity, a.k.a. the view from nowhere – shouldn’t exist, then the optimal course of action is to eliminate our specious preference for objectivity (different from factuality) itself, and replace it with honesty and a commitment to reason. I, for example, don’t blame people for their victimisation; I also subject an article exhorting agricultural workers to switch to organic farming to more scrutiny than I would an article about programmes to sensitise farmers about issues with pesticide overuse.

Limitations of the Finkbeiner test

This post was republished on The Wire on January 8, 2018.

The Finkbeiner test, named for science writer Ann Finkbeiner, was created to check whether a profile of a female scientist published by a mainstream news outlet was produced in the first place because its subject was a woman. It’s a good check to make when writing about a professional scientist’s work; if you’re going to write the piece because the subject’s a woman and not because you think her work is awesome, then you run the risk of presenting the woman as extraordinary for choosing to be a scientist. However, more than being a good check, it could also be too subtle an issue to expect everyone to be conscious about – or to abide by.

As The Life of Science initiative has repeatedly discussed, there are many systemic barriers for India’s women in science, all the way from each scientist having had few role models to admire growing up to not being able to stay in academia because institutional policies as well as facilities fall short in being able to retain them. And apart from working towards making these deficiencies known to more people, women have also been leading the fight to patch them once and for all. As a result, talking about successful women scientists without also discussing what needed to fall into place for them could ring hollow – whereas the Finkbeiner test seeks to eliminate just such supposedly miscellaneous information.

For example, a 2015 report by Ram Ramaswamy and Rohini Godbole and a 2016 article by Aashima Dogra and Nandita Jayaraj both stressed the need for affirmative action on part of the government so more women are retained in scientific pursuits at the higher levels. This means science journalism that focuses on a working woman scientist because she belongs to a particular gender and not on her scientific research at the outset becomes useful in the eyes of young scientists but also quickly fails the Finkbeiner test. Does this mean the piece becomes detrimental? I’d think not, especially because it would certainly serve the function of holding the people charged with instituting policy and infrastructural corrections accountable.

For another example, I’ve learned from several The Life of Science profiles that one reason many of the women who have become successful scientists with faculty-level positions were backed up by supportive families and partners. One profile in particular – of Mayurika Lahiri – stood out because it discussed her research as a cancer biologist as well as her achievement in setting up a full-fledged daycare centre in IISER Pune. However, the Finkbeiner test penalises an article on a woman scientist if it discusses her spouse’s occupation, her childcare arrangements or the fact that she could be a role model.

Two notes at this point. First: Some women might not like to be characterised in a way that the Finkbeiner test says they shouldn’t be characterised as. In such cases, the journalist must and will respect their choice. Second: To be fair to The Life of Science, the Finkbeiner test is intended only for mainstream publications and not specialist projects. At the same time, this caveat could come off as short-sighted because it aspires to make a stronger distinction between changes that remain to be effected for (India’s) women in science to have it as good as its men already do and the outcomes of those changes that have been implemented well. Persistence with the former results in the latter; the latter encourages the former to continue.

In countries where women receive more institutional support than they do in India, it’s possible to expect meaningful insights to arise out of applying the Finkbeiner test to all mainstream profiles of women in science. In other countries, the test could be altered such that,

  1. A discussion of women’s needs is treated on an equal footing with their science instead of having to ignore one or the other – This way, writers will have an opportunity to make sure their readers don’t take the pervasiveness of the conditions that helped women succeed for granted while also highlighting that their work in and of itself is good, and
  2. Profiles of male scientists include questions about what they’re doing to make science a non-problematic pursuit for people of other (or no) genders, if only to highlight that men often have a mission-critical role to play in this endeavour.

Featured image credit: bones64/pixabay.

A Nobel Prize award ceremony underway. Credit: nobelprize.org

Why do we cover the Nobel Prize announcements?

The Nobel Prizes are too big to fail. Even if they’ve become beset by a host of problems, such as:

  1. Long gap between invention/discovery and recognition,
  2. A large cash component given to old scientists,
  3. Limiting number of awardees to three,
  4. Not awarding prizes posthumously,
  5. Not awarding prizes to women, especially in the sciences, and
  6. Limiting laureates to those who had published in English or European languages*

… they have been able to carry over the momentum they accrued in the mid-20th century, as an identifier of important contributions, into the 21st century. The winner of a Nobel Prize gets his (it’s usually ‘his’) name added to a distinguished list, and has the attention of the world’s press turn towards him for 12-24 hours. The latter in particular is almost impossible to achieve otherwise. As a result, the Nobel Prizes, for all their shortcomings, still stand for a certain kind of recognition that is not easily attainable through other means.

Any other prize instituted today with the same shortcomings as the Nobel Prizes will struggle to be taken seriously (unless the cash component is overwhelmingly high). It is thanks to these qualities of its legacy that even those who write against the Nobel Prizes and their import can at best hope to fix the prize, and not have it cancelled. And this is also why people continue to lament problems #3 and #5 instead of neglecting the Nobel Prizes altogether.

I personally wish the Nobel Prizes stopped being important – but it’s a conflicted desire because of two reasons:

  1. It’s an opportunity – even only if it’s for one week of the year – to talk about pure science research instead of having to bother with what it’s good for, and still be read. Otherwise, there’s a high cost attached to ‘indulging’ in such articles.
  2. The Nobel Prizes are not going to drop in value among the people if only I abstain from covering them. Either all journalists have to stop giving a damn (they won’t) or the Nobel Committee itself will have to rethink the prizes (so far, they haven’t).

So if only I sit out and not write about who won which Nobel Prize for what, only I – rather, The Wire – loses out. I’d much rather make a bigger deal of homegrown awards like the S.S. Bhatnagar Prize, specialised prizes like the Wolf, the Abel and the Lasker, and the international – and more au courant – Breakthrough Prizes.

*I’m speaking only about the science prizes.

Credit: coldbrook/Flickr, CC BY 2.0

Some thoughts on the Mack/Dorigo Twitter exchange, and Zivkovic, Feyerabend, etc.

This exchange made me squirm:

(In case Dorigo deletes his tweets, screenshots here, here and here.)

If you didn’t know: Katherine Mack is a theoretical astrophysicist at Melbourne University and Tommaso Dorigo is an Italian particle physicist working at CERN. Mack’s Twitter feed is one of the best places to learn about astrophysics, and Dorigo’s blog is one of my preferred sources of information and analysis of LHC results. I consider them both very knowledgeable people. At least, I used to – until this short exchange on Twitter disabused me of the notion that they might be equally knowledgeable.

As my friend put it, Dorigo’s comment “makes it sound like being bi is a privilege” – especially since Mack goes on to detail the non-privileges being bisexual comes with. While I’m familiar with the issues surrounding gender and sexuality, I’m not entirely conversant with them, and yet even I know that Dorigo is being facile and refusing to engage substantively with the topic at hand. His response to Mack’s sharing the link is proof enough, conflating two attributes in a way that makes no sense:

I’m inclined to call this “Dorigo’s fall from my graces”. Some would argue that we ought to separate his technical expertise with his views on topics that seem to not directly relate to what made me pay attention to him in the first place. But I’m becoming increasingly wary of this line, particularly since allegations of sexual harassment were visited upon Woody Allen in 2014. While many hold that an appreciation of his films doesn’t require one to be okay what kind of a person he is, I disagree because the separation of professional achievements and personal conduct overlooks how one might enable the other, and together help establish structures of power and authority.

My example of choice with which to illustrate this is Bora Zivkovic, the former ‘All Father’ of Scientific American‘s famous network of blogs. His leadership as well as abilities as a communicator made young and aspiring writers flock to him for advice and favours. However, a string of allegations (of harassment and impropriety) emerged in 2013 that put paid to his job and, at least temporarily, his career. It was obvious at the time the scandal broke out that Zivkovic had abused his position of power to take advantage of trustful women and solicit crass things from them. When I first heard the news, I was devastated.

Now, science – rather, STEM – and science journalism already have a problem retaining women in their ranks. When they do, sexual abuse, harassment and sexism are rampant, often ensconced within organisational structures that struggle to remain cognisant of these issues. So when you embed men like Zivkovic and Dorigo – and, of course, Geoff Marcy – into these structures, you automatically infuse the structures with insensitivity, ignorance, etc., as well as increase the risk of women running into such men. And by paying attention to Dorigo – even when he’s talking about hadron-hadron collisions – I feel like I will be feeding his sense of relevance and legitimising his persistence as a scholar of note.

(Caveat: I’m keenly aware that mine could be a precarious position because it could displace a very large number of people from my self-aggrandising graces, but I choose to believe that there are still very many people who are good, who are aware, sensible and sensitive, who are not abusive. Katherine Mack is a living example; Dorigo would’ve been, too, if he’d had the good sense to apologise and back off.)

So where does Paul Feyerabend fit in?

From his Against Method (fourth edition, 2010; p. 169-170):

I have much sympathy with the view, formulated clearly and elegantly by Whorf (and anticipated by Bacon), that languages and the reaction patterns they involve are not merely instruments for describing events (facts, states of affairs), but that they are also shapers of events (facts, states of affairs), that their ‘grammar’ contains a cosmology, a comprehensive view of the world, of society, of the situation of man which influences thought, behaviour, perception. … Covert classifications (which, because of their subterranean nature, are ‘sensed rather than comprehended – awareness of [them] has an intuitive quality – which ‘are quite apt to be more rational than over ones’ and which may be very ‘subtle’ and not connected ‘with any grand dichotomy’) create ‘patterned resistances to widely divergent points of view’.

(Emphases in the original.) Our language influences the weltanschauung we build together. While Feyerabend may have written his words in relation to his idea of incommensurability in the philosophy of science, their implications are evident in many spheres of human endeavour. For example, consider product advertisement: a brand identity is an intangible thing, an emotion trapped within a cage of words, yet it is built and projected through tangible things like design and marketing all embodying that emotion.

Similarly, involving this or that scientist in a conversation is to include a certain point of view that – even in the presence of robust safeguards – suggests not an endorsement but definitely a willingness to ignore something that may not always be ignorable.

Featured image credit: coldbrook/Flickr, CC BY 2.0.