For much of the modern, advertising-driven Internet, attention is money. Whether a publisher, search engine, or social platform, the business model is fundamentally the same: The attention of users is aggregated and then sold to advertisers who want to bring a message to that audience.
To that end, supplying attention itself can be an act of complicity in the unethical actions of a platform. The mere act of choosing to look at something online generates real value for a company, materially helping to support its staff, its content, and the social interactions that a platform plays host to. This is why a website like Do Not Link exists: It promises a way to share a link from a website without boosting that site’s standing in search rankings.
Like environmental, labor, or civil rights activists of an earlier decade, one response has been to boycott objectionable platforms. At first blush, this comparison seems very straightforward: refusal to click or link to a site denies revenue to ad-driven companies the same way that refusing to purchase goods denies revenue to retail-driven companies. Like earlier campaigns, these choices are in part about influencing the company to behave differently, but also in part a personal choice to avoid becoming part of the ethical wrongs being perpetuated by that company. It’s a way of voting with clicks instead of cash.
But while the motivations may be similar, the Internet distorts the traditional dynamics of the boycott in two important ways.
For one, the choice to refuse to buy sweatshop-produced goods influences demand: people avoiding giving an unethical actor money for their products. In contrast, the boycott of attention influences supply: People refuse to give platforms the raw material that they can in turn sell to advertisers for money.
This fact matters because it places the protestor on the wrong side of the Internet business equation, where advertisers are far outnumbered by the number of ldquo;eyeballs” they bid for. The loss of an average user matters such a minuscule amount financially that the importance of the boycott narrows to being symbolic and personal, rather than striking directly at a bottom line.
Secondly, making attention the unit of valuable, measurable inventory makes the nature of a boycott complicated in a way that doesn’t come up when refusing to hand over hard currency for a product.
For example, one might never engage with racist activity on a platform like Reddit, and might indeed even use the same platform to protest against this kind of activity. However, the very fact of being a user increases the overall stock of attention commanded by a site, which in turn provides real money to the very forum that facilitates the behavior some users otherwise oppose. In contrast to a world where a boycott only requires that an individual stop buying from a given company, an attention boycott forecloses nearly any activity having to do with the site being challenged.
So, what does it mean to be a conscientious consumer in such an economy? Is it even sensible to try to surmount these challenges?
Two recent high-profile examples show the shape of an emerging notion of what we might call ethical attention. More than simply symbolic gestures, the refusal to link or click actually expresses a deeper vision of the role norms play in the evolution of the Internet, and more importantly, new ethical expectations for individuals and platforms on the web.
* * *
First, a representative example. Last month, the editorial team of Gawker made the choice to publish a story which exposed the alleged attempt by a CFO of a prominent magazine company to hire a gay porn star for sex. Glenn Greenwald of The Intercept chimed in soon after to pick apart the justifications that Gawker’s editorial leadership provided for the decision to publish.
Greenwald’s piece also provides an explicit refusal to provide any links to the controversial post. As he wrote simply, “I don’t want to reward them or contribute in any way to this disgrace by linking to it: Google it if you must.”
This is just one prominent example, rather than something entirely new. Many others followed a similar path in refusing to link to the Gawker story. Refusals to link come up relatively frequently, and in all sorts of other contexts, too.
Choosing not to link is curious in a universe where the availability of search engines make content—particularly content generating lots of attention—quickly and easily findable. It isn’t like the refusal to link actually prevented users from accessing the content if they wanted to. The Gawker article generated a huge number of pageviews and trended on a number of social platforms, regardless of Greenwald’s choice not to link.
The choice not to link is therefore a personal moral act: It invokes an individual responsibility around making content accessible to others online. The economics of advertising are such that linking provides a frictionless channel for an audience’s attention (read: money) to reach content. The web of information stitched together by an individual as they browse and publish across the Internet is also implicitly a web of support for the content being linked to.
This shuffles up our traditional notions of what it means to link. Linking is tangled up with our concepts of proof and good argumentation online. One links to something else in order to provide a citation that backs up a point—that’s how I’m using links in this very article, for instance. The often-heard call of “citation needed” on Wikipedia echoes much of the same functionality.
Choosing not to link in that context represents a belief that the ethical duties around linking will sometimes outweigh the need to use linking to facilitate discourse and debate online. In some ways, it implies that the latter use is the lesser necessary as the ability to find information has grown enormously from the early days of the web.
Let’s take another representative example. In the same month as the controversy above, Gawker published a personal reflection by Reginald Braithwaite, a black user of Reddit, explaining his reasons for departing the platform.
The core of the post focuses on the fact that Reddit has become, by some accounts, home to the biggest white supremacist community on the web. To that end, the continued participation by Braithwaite, even in parts of the site that didn’t engage in racist behavior, provided material support to racism. As he wrote succinctly, “every page view turns into some fraction of a dollar that powers a server that hosts hate.”
Again, this is just one of a wide range of instances where people have refused to personally participate in a platform because of its policies. The protest here is useful, though, because it is distinct in substance from the earlier Gawker example. Gawker is a publisher of content, and in doing so its writers and editors act as agents of the organization. Refusing to link to them makes sense in a direct way because a user refuses to provide attention and advertising revenue to the content created by the organization.
Reddit introduces a wrinkle precisely because it is a user-generated platform. For one, no one (yet) has claimed that the individuals contributing racist or sexist content are affiliated with Reddit as employees of the company. Instead, part of Braithwaite’s protest is simply that the platform plays host to a sizable, and perhaps one of the largest, community of racists online.
But it is more than simply hosting. If it were just this, the scope of the moral commitment could easily become enormous. Facebook is a platform with over a billion users. Some of the content and social activity on Facebook is racist and sexist in the same way as Reddit. Does that make the use of Facebook ethically suspect?
This gets even more fraught if you dive deeper into the the structure of the web. Amazon operates S3, a hosting service that serves as the backend infrastructure for a huge portion of sites online. Arguably, providing attention to one ad-based platform hosted on S3 also indirectly provides support for infrastructure that likely plays “landlord” to other platforms that are home to racist hate. It cannot be that one drop of objectionable behavior renders a whole platform ethically problematic.
What appears to be relevant in the Reddit case is the notion that the company has exercised a kind of negligence towards the organic behavior emerging on the platform. While Reddit does not create the content or even promote the content, the failure to act makes continued use of the platform tantamount to a moral complicity in the emergent behavior of other users. Platforms, in short, are deemed morally responsible for the consequences, even unintentional, of the spaces they create online.
* * *
There is nothing absolute about “the Internet.” From the history of the web to a look at the structure of the network across countries, it is obvious that the configuration of networks, code, and people that make up what we call the Internet evolves continuously.
To date, one notion around this has been what David Weinberger calls the “Argument from Architecture,” the idea that values embedded in technical features of the web generate certain norms in society. These values, as he writes, include “open access to information, the democratic and permission-free ability to read and to post, an open market of ideas and businesses, and … bottom-up collaboration among equals.”
These norms are argued to have follow-on effects in society at large, and are also supposed to produce interventions that themselves sustain this notion of what the Internet is. Section 230 of the Communications Decency Act, for example, famously grants platforms a legal immunity in ways that forward these “architectural” values.
Both the refusal to link and the refusal to click invert the idea that the technical aspects of the web necessitate certain norms. Instead, they seem to affirm the belief that the ethical norms of users and platforms can and must constrain the open structure of the web, rather than the other way around. In the very least, it suggests that the ubiquitous economics of advertising currently in effect demand a different relationship between users, ethics, and platforms.
Refusing to link affirms the idea that individual users have personal moral duties to constrain open access to information when to not do so would furnish financial benefit to unethical behavior. In that respect, it rejects the idea that all content is made equal for the purposes of citation online, and adopts a view that how content is produced is ethically important.
Refusing to click and participate on a negligent platform affirms a moral principle that platforms have responsibilities to limit an open and generative market of ideas where to not do so would produce social harms. In some ways, refusing to click discards the entire idea that a platform may be fully “neutral” in its design, and accords it with influence and obligations to actively govern the communities that emerge within it.
Refusals to link and refusals to click are independent actions taken around individual issues, rather than part of a common movement around ethical attention. But they are influencing the creation of tools that put these moral intuitions into action. Just as Do Not Link allows the sharing of information without triggering the follow-on benefits of an increased ranking in search results, The Block Bot aggregates users engaging in harassment and allows for their easy blocking on Twitter.
Norms might become embodied in how we use existing tools, as well. The ad blocker, currently marketed as a way to avoid the distractions and privacy intrusions of advertising, takes on a new life as an ethical act in a world of conscientious clicks. One might imagine a future mutation of this tool which only permits ads on sites that meet certain ethical protocols defined by a user.
It is unclear at the moment to what degree practices of ethical attention are widespread domestically or internationally. Regardless, as they proliferate, these small acts contain within them seeds of a radically different and much larger vision of the priorities and structures that should shape social life on the web.
We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.
