Communicating science without scientists: what Chinese social media reveals about who gets to speak for science

22 July 2026

By Anfan Chen

An influencer with two million followers spends fifty seconds explaining why blue light delays the onset of sleep. Beneath the video, a purchase link for melatonin sits on screen for all fifty seconds. The explanation and the sale are not two acts stitched together. They are one act, occupying the same screen and the same seconds.

Nothing about this is unusual, and my question is not whether it is good or bad. My question is who this person is, and why they, rather than a sleep researcher, are the one explaining melatonin to two million people.

The answer is not that Chinese audiences are uninterested in scientists, nor that scientists are silenced. It is that the space where the public meets science has been platformised much faster than it has been professionalised, and the people best positioned to occupy it are not the people our field has spent thirty years trying to train.

Why the scientists are not there

Hepeng Jia’s diagnosis of Chinese science communication, “more engagement but less participation”, describes an enormous state-supported science communication (科普, meaning science popularisation in Chinese) apparatus that produces a great deal of outreach but very little dialogue. On social media, this shows up as scientists being, in Jia and colleagues’ earlier phrase, encountered but not engaged: even the most active online communicators among them favoured one-way knowledge dissemination over dialogue, and most working scientists were not on these platforms at all.

A national survey of 8,533 scientists, fielded through the China Association for Science and Technology, gives some sense of why. Chinese scientists reported clearly higher intentions to do outreach through legacy media (M = 3.51) than through social media (M = 3.01). The strongest predictors of doing it on social media anyway were past participation and personal norms, an internalised feeling of moral obligation. Perceived external rewards did not drive the behaviour. They worked against it: extrinsic rewards were negatively associated with personal norms, potentially eroding the moral motivation that drives participation. Believing that one’s colleagues are out there posting also predicted social-media intentions negatively.

The reputational side is sharper still. In a companion study on the Sagan effect, around 30% of the same scientists agreed that outreach had damaged their reputation or their career progression, and those experiences depressed future intentions indirectly, by installing a felt obligation not to participate. This sits inside a public culture in which experts are routinely mocked as zhuanjia (砖家 in Chinese), a homophone replacing “expert” with “brick”, and in which official backing subtracts from an expert’s standing rather than adding to it. Joy Y. Zhang named this the credibility paradox a decade ago: the scientific practitioners she interviewed found that an absence of visible institutional endorsement won them more public credibility, not less. More recent work on expert trust on Weibo suggests the gap has widened.

So, the credentialed supply is thin, institutionally mediated, self-motivated where it exists at all, and exposed. That is the vacancy.

What filled it

What filled it is not a public sphere. It is a market. Chinese platforms industrialised creator work early: Douyin and TikTok platformised in parallel, but Douyin folded shoppable links and livestream commerce into ordinary content production far more deeply, and it instructs creators explicitly in how to align their output with those commercial priorities. The pull is not limited to amateurs. Several of the scientist-communicators interviewed by Dapeng Wang, Hepeng Jia and colleagues already reported substantial income from platform advertising-revenue sharing and readers’ tips. Even professional science-communication start-ups in China follow the metrics, shaping topics and formats around visibility.

In a study we are completing, we mapped 219 active science influencers and 6,769 of their Douyin videos across three months. Only about a third carried any institutional endorsement — the visible marker of scientific authority is a minority feature of this population, not its defining one. About a quarter of the influencers posted shoppable videos, and 15% of all videos carried a merchandise link.

Three findings matter for how we think about this. First, audiences do notice: commercial videos were the least engaged content their own authors produced, ranking below everything else on likes, comments, shares and saves. Second, selling is a lottery. Half of all commercial videos recorded no sales at all, and the top 5% accounted for 92% of sales and 96% of revenue. Third, and this is the finding I keep returning to, fit does not pay. Social-science influencers, promoting the goods least related to their content, earned the most; natural-science influencers had the best content–merchandise fit and no commercial advantage. Where fit is low and returns are high, the science is not explaining the product. It is supplying the setting in which the product appears.

Read together, these patterns say that who speaks for science on this platform, and what gets monetised, depend far less on scientific substance than on an account’s position in the platform’s traffic and commercial systems.

What follows

At the end of 2024, China revised its Science Popularization Law for the first time in 22 years, adding a chapter on science popularization personnel and strengthening incentives, honours and evaluation. It is a commitment aimed at the supply side, at persuading credentialed scientists to produce more. Our survey data suggest that the instruments most readily available to any state or institution, namely extrinsic rewards, are precisely those that undermine the norms doing the real work. And the deeper problem is not a shortage of authoritative content to inject into the feed. It is that the feed already has occupants, operating on terms no content strategy addresses.

Three things follow, and none of them is specific to China. Design recognition rather than payment: reward structures should activate professional identity, not substitute for it. Study the account rather than the post: commercialisation on these platforms is a portfolio arrangement, which is why disclosure regimes that treat each commercial post as an equal unit of risk are badly matched to a distribution where a handful of videos carry almost all the money. And stop treating Chinese platforms as an exotic comparison case. Shoppable video, creator-economy incentives and platform-run debunking are arriving everywhere; China simply got there first.

The public will keep meeting science somewhere. The only open question is who is standing there when they arrive.

Anfan Chen
Research Assistant Professor in the Department of Communication Studies at Hong Kong Baptist University, where he serves as Associate Director of the AI Media Centre, Executive Editor of HKBU Fact Check and Programme Director of the Master of Social Sciences (MSocSc) in Media Management. His research focuses on science communication, misinformation and Chinese social media platforms.

Reading suggestion

AI-use declaration
This post was written by the author. Anthropic’s Claude was used for proofreading only; the argument, the data, the examples and the final wording are the author’s own. No part of the content was generated as finished text by AI.

The views and opinions expressed in the PCST Insights series are those of the author(s) and do not necessarily reflect the views, positions, or policies of the PCST Network, the Scientific Committee, or its members as a whole. This series is intended to foster thoughtful discussion and showcase a diversity of perspectives from across the global science communication community.

Previous PCST Insights

➔ Read all our PCST Insights

Become a member

We invite people from round the world to our join our science communication network. Whether you are new to the field or an experienced professional, you are welcome.

➔ Find out more and join us

Join the discussion

The discussion list is free for everyone to join and has an international audience of over 2000 subscribers.

➔ Find out more and subscribe

Contact

The Network for the Public Communication of Science and Technology Incorporated (PCST Network) is an organisation that promotes discussion on the theory and practice of communicating science, and of public discourses about science and technology and their role in society.

Email: admin@pcst.network
Web: www.pcst.network

Incorporated in the Australian Capital Territory, Australia. Registered number AO5352.

Connect

PCST is the an active and influential international academic and professional organisation in the field of science communication.

➔ Become a member

Our email discussion list is free for everyone to join and has an international audience of over 2000 subscribers.

➔ Join the discussion list

PCST Network has co-operation agreements with the following international organisations:

EUSEA
European Science Engagement Association

ICA
ICA Science Communication Interest Group

IECA
International Environmental Communication Association

RedPOP
Latin American and Caribbean Network for the Popularisation of Science and Technology