Weak Ties

Weak ties are casual interactions with people you barely know: the barista, the security guard on the desk, the dog-walker you pass at the same hour. Two different claims travel under the name and they come from two different literatures. The term is Mark Granovetter's, from 1973, and his subject was information flow through social networks, specifically how people find jobs. The wellbeing claim, that these low-stakes exchanges carry measurable benefits for mood and belonging, comes from Gillian Sandstrom and Elizabeth Dunn in 2014. Keeping the two apart is the whole point of this concept.
What it is
Granovetter's argument is structural. Acquaintances outperform close friends as a route to a job because they reach into networks your own does not cover, so the information a person most needs arrives through the connections they invest in least. That is a claim about networks, not about how a chat with a barista makes anyone feel. It has held up well and has been tested at scale: a large field experiment on a professional networking platform, reported by Karthik Rajkumar and colleagues in Science in 2022, found that moderately weak ties were the most useful for job mobility.
The wellbeing line is separate. Sandstrom and Dunn found that students and adults felt better, and reported more of a sense of belonging, on days when they interacted with more weak ties, with a companion study running the same question in a coffee shop. These are small, largely correlational, diary-based studies. Later work on talking to strangers points the same way, and no large preregistered replication of the wellbeing claim was located.
In effect
The everyday form is easy to recognise. Saying good morning to the same person at the same bus stop for two years without ever learning their name, where nothing is disclosed, nothing is owed, and the exchange still registers.
Zelana Montminy's Finding Focus is where the concept reaches this library, and the argument there is that these interactions are the ones people cut first when attention is scarce, because each is individually skippable. Her sourcing is worth recording. The prose puts weak ties in quotation marks, which signals the term belongs to somebody, and credits nobody, with no name, year or source anywhere on the page. The wellbeing work she is drawing on is correctly cited in her endnotes, and it is the right citation for the claim she makes. So the book borrows the name from one literature while sourcing the evidence to another, which is a fair borrowing and should be labelled when the term is used.
What it does not say
It does not say that acquaintances make people happy because acquaintances find people jobs. Those are two findings about two different things, and the second does not inherit the first's stronger support.
It does not establish a size for the wellbeing effect. The studies print no effect anyone should quote, the designs are diary-based and correlational, and no large preregistered test was found.
It does not say weak ties substitute for close relationships. The literature treats them as an additional layer, and nothing here supports trading one for the other.
Sources
- Granovetter, M. S. (1973). "The Strength of Weak Ties." American Journal of Sociology, 78(6), 1360-1380. The term, and a network argument about information and job-finding.
- Rajkumar, K., et al. (2022). Science. A large field experiment on a professional networking platform, supporting the network claim at scale.
- Sandstrom, G. M., & Dunn, E. W. (2014). "Social interactions and well-being: The surprising power of weak ties." Personality and Social Psychology Bulletin, 40(7), 910-922. The wellbeing claim. A companion paper by the same authors in Social Psychological and Personality Science (2014) is the barista study. Small, largely correlational, diary-based.
- No large preregistered replication of the wellbeing claim was located.
- Finding Focus, Zelana Montminy. The prose uses "weak ties" in quotation marks and names nobody, with no year, journal or sample size. The endnote correctly cites Sandstrom and Dunn. Granovetter is our identification and is not in the book.
- Evidence status mixed, following the verification verdict of 2026-09-25, which supersedes the earlier primary-cited reading in the vault note.