The National Academy of Sciences just posted a paper titled “Online Social Integration Is Associated With Reduced Mortality Risk.” I gave a short comment on the paper’s findings to New York Times reporter Jonah Bromwich, but I wanted to expand on those thoughts here. This paper is of interest to me because, in evaluating whether digital connection has similar positive health effects as we know social connection in general has always had, it touches on whether social life should be seen as split into online and offline forms. That question is foundational to my work — and part of the premise of this publication.

Despite some heartening conclusions about the reality of digital connection, this paper unfortunately begins from an assumption that social activity in digital networks is separate and not an inherent part of the “real world.” This presupposition affects everything from the paper’s research questions to its methodology to the framing of its conclusions in dualist language. And ultimately the dualism undermines the paper’s claims.

The substance of the paper looks at the mortality rate for Facebook users in California over two years, and whether different ways of using the site were associated with being more or less likely to die in that timespan. The study could not include income as a variable, which hampers it methodologically, but the authors do a good job explaining that their finding — that people who don’t use Facebook die at a higher rate — is accordingly compromised: They acknowledge that it’s probably the case that those who have more money use Facebook more and also die at a lower rate, but they can’t know for sure without income data for their sample.

The authors’ further claim that Facebook users who post more photos than text statuses die at a lower rate suffers from the same missing income data (though the paper does not acknowledge it this time). The authors suggest that photo-sharing might be more related to face-to-face interaction, which in turn might be related to better mortality outcomes. I like the creativity of this interpretation! But their assessment that text statuses are “online only” whereas sharing photos is a proxy for being with people “in person” is a bit of a stretch relative to the simpler explanation that the diminished mortality has more to do with income. There is some evidence that photo-based social media use is related to income. Phone manufacturers have long marketed new or more expensive phones on the basis of their cameras, establishing a link between photo-taking, status, and wealth, and Pew found that higher-income teens are more likely to use photo-based networks. So people who post more photos perhaps have more money, and those with more money die at a lesser rate as a population over a given time, all else equal.

But more problematic than the data’s shortcomings may be the paper’s conceptual issues. Though the study provides much evidence against digital dualism, it unfortunately couches all this within dualist presuppositions and language. “Digital dualism” is a phrase I coined to describe this view that digital connection makes for some other, virtual, or cyber space that exists separately from the “real world,” as if one is always switching back and forth between online and offline worlds. The way this paper speaks of “online social media” versus the “real world” is precisely such a dualist framework. So deeply digital dualist are our presuppositions that even challenging this dualism is spoken in dualist terms.

What this suggests, for me, is that saying things are “on-” or “off-” line is a mistake to begin with. Digital dualists overcomplicate their language and logic to make sense of a model of two worlds, one on and one offline, in the face of evidence that suggests otherwise. It reminds me of the complicated planetary “epicycles” that ancient astronomers devised to reconcile their model that placed the Earth at the center of the solar system with the recorded observations of planets’ movements.

The paper’s claim that “online social media” is related to the “real world” is anachronistic, positing as discovery something that should be taken for granted in the first place. When this paper’s authors find that “online social life predicts differences in mortality rates, confirming that it does so in the same way that offline social life does,” they are using dualist language to describe why their own dualism makes little sense. Congrats, you blew down your own straw-ontology.

The only way these conclusions are surprising is if you assumed that social media is a separate world, and that longstanding social processes like the link between status and health could be thrown out in this newfangled cyber space.

The point of social science research on social media shouldn’t be to continuously rediscover the fact that communicating through a new medium still takes place in reality. What is important is that social relations transpire through many flavors of information (text, voice) and these all have varying impacts, based on, among other things, the presence of screens and the affordances of different apps for different people in different times and places. But the actual reality of these modes of communication on our lives should be understood. Digital connection, for all its experiential novelty, is part of this one social reality.

Acknowledging this from the outset would save us all the conceptual and linguistic backflips researchers make in trying to demonstrate that some virtual world interacts with the real world. For example, the paper at hand finds that accepting (as opposed to asking for) friend requests is associated with positive social outcomes. But talking about how being popular online has offline health benefits or how being popular offline shows up online and has health benefits is just a distraction from the simpler conclusion: those who are more strongly connected (which is an infinitely complex matter of screens and bodies) tend to have better social outcomes.

Long-established social processes can’t be totally dismissed in a world with screens because those screens are part of this world. Those researching social media — as well as those people who are designing them — would do well to acknowledge that social media are part of the real world, not some new cyber space with a totally new set of processes to be uncovered. There is no on- or off-line, digitality is continuous with social reality and is conditioned by its history. We should design networks with an understanding of and respect for long-standing social processes, even ones we don’t like, rather than trying to “disrupt” them.