Being alone is often treated as a single condition: either another person is present or nobody is. A 2024 study suggests that this is too blunt. Among 888 adults, the form of solitude that left people entirely inaccessible and without media was rated as less restorative than a lighter version in which they were not communicating but could still be reached, read a book, listen to music or browse online.
This is one study, not settled consensus. It was a cross-sectional survey of what people expected different situations to feel like, not an experiment that sent participants into isolation and measured how well they recovered. The headline’s phrase “total isolation” refers to the researchers’ most complete form of short-term solitude. It does not refer to chronic social isolation, solitary confinement or a clinical intervention.
The result is intriguing precisely because it challenges an intuitive picture of rest. If conversation and demands drain us, it seems reasonable to think that removing every remaining channel to the outside world should produce the deepest recovery. In this dataset, however, restoration did not rise with completeness. It declined.
The study asked what counts as solitude
Smartphones complicate the familiar definition of solitude as simply being physically alone. Someone can sit alone in a room while exchanging messages with several friends. Another person can choose not to communicate while sitting among strangers in a café. Researchers Morgan Quinn Ross and Scott W. Campbell developed a noncommunication framework to capture that distinction: solitude depends not only on physical separation, but also on whether communication is taking place or remains possible.
For their study in PLOS ONE, the researchers recruited 1,469 US adults through the online research platform CloudResearch in August 2023. They excluded 429 people who failed an attention check, 37 who asked for their data to be removed and 115 whose answers about how often they experienced solitude were starkly inconsistent. That left 888 participants.
The final sample had a mean age of 61.9, with a standard deviation of 14.9 years. It included 359 men, 525 women and four participants who described another gender identity. That relatively old sample matters. The findings may be more informative about older adults’ expectations than about the population as a whole.
Four shades, from non-interaction to noncommunication
Participants considered four nested situations. The least complete, called “base solitude”, meant choosing not to communicate in person or through technology. People in this condition could still be accessible to others and could use media. A second shade added inaccessibility, while a third removed media. “Total solitude” combined both restrictions: no interaction, no potential interaction and no engagement with media.
In other words, “being alone while still loosely connected to the world” is a useful shorthand for base solitude, but it is not a licence to imagine an active group chat. Communication was still paused. What remained was the possibility of contact and access to activities such as books, music, television or websites.
The researchers always presented the base condition first and total solitude last. The two middle conditions were randomised. Participants rated how likely each would be to offer restoration and a sense of connection. The nesting is conceptually useful, although some participants said the definitions were difficult to distinguish, which the authors acknowledge as a limitation.
Total solitude ranked lowest for restoration
Restoration was measured with seven self-report items. Participants rated, on a seven-point scale, how likely each form of solitude was to help them recover, including whether it would “renew energy”. Base solitude received a mean restoration score of 4.18. The mean was 4.09 for base solitude plus inaccessibility, 4.02 for solitude without media and 3.85 for total solitude.
The absolute differences were modest, but the direction was clear enough to contradict the researchers’ prediction that total solitude would be the most restorative. The statistical comparison across conditions was significant. Removing every remaining thread of connection did not appear to create a larger anticipated refuge.
Relatedness generally fell as more connection was removed, from a mean of 4.20 in base solitude to 3.60 without media, 3.19 when inaccessible and 3.34 in total solitude. That last reversal was unexpected. When restoration and relatedness scores were combined, base solitude ranked highest at 8.37, followed by no media at 7.62, inaccessibility at 7.27 and total solitude at 7.18. The preregistration, study materials and data are available through the researchers’ Open Science Framework project.
This was a survey of expectations, not an intervention
The crucial restraint is that these scores represent perceived likelihood of restoration. They are not measured changes in fatigue, cortisol, attention or mood after participants spent time in four controlled environments. Nobody was randomly assigned to switch off every communication channel, and participants were not prompted to recall one specific recent episode. Cross-sectional data cannot show that removing access or media caused lower restoration.
The categories also bundled together experiences that may work very differently. Reading a novel and scrolling through upsetting news both counted as media. Remaining reachable by a trusted friend and waiting for a demanding work message both counted as accessibility. The design can compare broad shades, but it cannot tell us which book, screen, person or notification helped or hindered.
There is another wrinkle. In an exploratory analysis limited to participants who said they had experienced chosen solitude, restoration did not differ significantly across the four shades. That does not erase the main result, but it makes a universal recommendation harder to defend. The data support a question worth testing experimentally, not a rule that complete disconnection is always inferior.
Choice, activity and beliefs change the experience
The reason someone is alone may matter as much as the number of open communication channels. Research behind the Motivation for Solitude Scale distinguishes self-determined solitude, such as choosing quiet because it feels valuable, from non-self-determined solitude driven by anxiety, rejection or a wish to escape an uncomfortable social setting.
A separate three-week diary study of 178 adults in the UK and US produced 2,967 daily reports. It found that autonomy during solitude predicted more peaceful affect and greater satisfaction with the day, while feeling competent was associated with less loneliness and higher day satisfaction in the published analysis. A person who chooses an hour with a book is not necessarily having the same psychological experience as someone who feels excluded for an evening.
Beliefs can shape the experience, too. A later programme of research found that seeing solitude as beneficial rather than harmful was associated with less loneliness, and that media messages could move those beliefs in either direction across five studies. This complements a broader conclusion from research on chosen solitude and the quality of one’s own company: time alone is neither automatically restorative nor automatically harmful.
A softer boundary may be the useful insight
The 888-person survey does not prove that everybody should keep a phone switched on. It does not show that a silent walk, a closed door or an offline afternoon cannot be deeply restorative. It also did not compare a forest with a busy café, or ten quiet minutes with a weekend retreat. Completeness and duration are different dimensions.
What it offers is a more precise way to think about a break from social demand. Solitude may restore us through more than absence. Agency, a chosen activity and the knowledge that contact remains possible may preserve something useful while active conversation pauses. For some people, the relief comes from being unavailable. For others, being unreachable may introduce a different kind of tension.
The practical question is therefore more specific than “Do I need to be alone?” It is “What kind of aloneness would help me recover today?” That might mean putting messages aside while reading in a public place, taking a walk with a podcast, or choosing complete quiet for a while. The study cannot select among those options for an individual. Its valuable suggestion is that solitude has shades, and the version with the hardest boundary is not automatically the one that restores us best.