Social Media and Mental Health: An Older Diagnosis

Chapter 12 is five lines about too much input. "The five colors the human eye will blind. The five notes the human ear will rend. The five tastes the human mouth offend. Racing and hunting will human hearts turn mad, Objects of prize make human conduct bad" (chapter 12, Paul Carus, 1898).

It was written about court luxuries — dyed silks, court music, banquets, the hunt. The claim underneath is not that colour and music are bad. It is that intense stimulation does not sharpen a sense. It dulls it. Enough of it and the eye stops seeing.

What the Research Actually Shows

It is worth being accurate here, because this subject attracts overstatement in both directions.

The large correlational literature finds associations between heavy social media use and depression, anxiety, and loneliness that are consistent but small — meta-analytic summaries put them in the region of r = .12 for depressive symptoms, .10 for anxiety, .15 for loneliness. Those are modest numbers. Most of that work is also cross-sectional, which cannot settle direction: feeling low may drive scrolling at least as much as scrolling drives feeling low. Anyone telling you the science is settled is ahead of the evidence.

The experiments are the more useful part, because they can speak to cause. Hunt and colleagues (2018, Journal of Social and Clinical Psychology) had 143 undergraduates cap Facebook, Instagram and Snapchat at ten minutes per platform per day for three weeks; that group showed significant reductions in loneliness and depression against a use-as-usual control. Allcott and colleagues (2020, American Economic Review) paid 2,743 people to deactivate Facebook for four weeks: subjective wellbeing rose, offline time with family and friends rose, political polarization fell — and factual news knowledge fell too. Many of them used the platform less afterwards, without being asked to.

So the honest summary is narrower than a headline and more useful: the average effect across everyone is small, and for people using a great deal, cutting back tends to help.

The Mechanism Has a Name

Research on why points repeatedly at one process: upward social comparison. Studies tracking daily use find that comparison carries much of the link between time spent and how people feel, and most of that work finds upward comparisons lowering mood and self-esteem — though not uniformly, since some content produces something closer to inspiration.

Chapter 24 describes both ends of that process without needing the vocabulary. "By displaying oneself one does not shine. By self-approbation one is not esteemed. In self-praise there is no merit. He who exalts himself does not stand high" (chapter 24, Walter Gorn Old, 1904).

A feed runs both halves of that at once and at scale. One side of the glass is display; the other is comparison. And the comparison is structurally uneven — what is available for comparison is an edited exterior, set against an interior nobody else can see. Chapter 24 says self-display does not produce the standing it reaches for. The research says being on the receiving end of everyone else's does measurable work on mood.

From the Tao

The five colors the human eye will blind. — Tao Te Ching, chapter 12 (Paul Carus, 1898)

Dulling, Not Enjoying

Chapter 12's claim is about capacity rather than morality, and that is what makes it land now. Intense input does not add pleasure indefinitely. It raises the level at which anything registers at all.

It is a common enough report: after a long stretch of scrolling, ordinary things — a book, a walk, a conversation with no cuts in it — feel slow rather than restful. Nothing has been lost from those things. The floor moved. That is what the chapter means by blinding: not that the eye stops working, but that the ordinary stops arriving.

What the Chapter Prescribes

It ends with the remedy in one line: "the holy man attends to the inner and not to the outer. He abandons the latter and chooses the former."

That is not a call to renounce the world, and it is not a demand to delete anything. It relocates the reference point. The question stops being how a thing compares and becomes whether it is any good from the inside — which is, incidentally, the only question a feed cannot answer for you.

The trials point the same way, and it is worth noticing how unambitious they were. Nobody in the Hunt study quit; they capped. The Allcott participants came back. Neither intervention required renunciation, and both moved the outcome.

Not a Case for Deleting Everything

The same research that found wellbeing rising off Facebook also found news knowledge falling, and participants who still put real value on the platform. There are things in there worth having: people you would otherwise lose, work that depends on it, whole friendships conducted in a feed.

Chapter 12 is not asking for a bare room. It is drawing a line between input you chose and input handed to you, and pointing out that only one of those is under your management. Most of the practical difference lies there rather than in deleting an account.