Our lives are lived day by day, hour by hour, and minute by minute. This fine-grained, equally spaced resolution of time means we do not always perceive changes in the small, everyday things around us, like the slow deterioration of the road we drive on every day. Sometimes we need to take a step back and examine them from a more distant perspective, for example, when unexpectedly hitting a newly formed pothole.
Such an “unexpected pothole” was recently afforded to us when analyzing several data sets jointly within the scope of a registered replication report (Tidwell et al., 2025). These data feature estimates of daily spoken words for 2,197 participants based on a careful analysis of passively sampled ambient audio recordings of their daily lives. The participants, ranging in age from 10 to 94, were part of 22 studies conducted over the span of 14 years and resided (mostly) in the United States as well as Mexico, Australia, and Europe. The report yielded an estimate for the average number of spoken words per day of 12,792, an estimate noticeably lower than what a prior study from 2007 yielded using the same method (M = 15,959; Mehl et al., 2007). After verifying that this was not a mathematical error, we decided to put the hypothesis that we had lost words over time to the test.
By relating the estimates of participants’ daily spoken words to the year their data were collected, we discovered that for each year between 2005 and 2019 an average of 338 fewer words were spoken per day, b = −338, 95% credible interval (CrI) = [−652, −25], as estimated using Bayesian multilevel modeling. At first glance, 300 or so words seems like a small, insignificant loss, as if it would not make much of a difference. And indeed, this finding ultimately was buried in a publication full of seemingly more interesting, larger, and more controversial findings.
But, as trivial as 338 words a day may feel, the loss of these daily spoken words unavoidably adds up. It means that, each year, we speak more than 120,000 words fewer than in the previous year. From 2005 to 2019, the reduction in the estimated number of words spoken per day was about 28%. This loss of words reflects real spoken conversations, big ones and small ones, that we stopped having with others. While putting a number to the loss, there is much about those lost conversations that these data cannot answer. Were they lost with friends, or family, or with strangers? Were they lost equally for everyone, or just for a select few? What were those conversations about? And, more consequentially, can we quantify the individual and societal costs of lost spoken conversations?
Although the time period of our data collection coincides with the rise in digital communication, such as texting, emailing, or social media, it is currently unclear whether digital conversations contribute to, and can make up for, a loss of spoken words. Considering this possibility, it would be important to better understand whether and how typed and spoken conversations are qualitatively different and how digital conversations outside of our in-person social networks contribute to loneliness or can help counteract it. If pressed for an opinion, we, as psychologists, would argue that spoken, “live” conversations involve different social and cognitive processes than typed conversations and that such live conversations can elicit benefits in well-being that apps and social media cannot readily replace.
One factor that might be informative when considering the role of technology in the loss of spoken conversation is age. Younger individuals are more likely to use technologies such as texting or social media to communicate than are older adults, and, if spoken words become typed, younger adults should lose more words than do older individuals. To test this possibility, we split our sample roughly in half at the developmental threshold of 25 years of age (separating adolescence and emerging adulthood from early, middle, and older adulthood) and reran our analysis within each age group. We found that participants below the age of 25 (n = 980) lost 451 words, 80% CrI = [−825, −89], for each year that passed since 2005, whereas participants above the age of 25 (n = 1,146) lost 314 words, 80% CrI = [−579, −41], per year (Fig. 1). Thus, younger participants lost about 44% more words than did older participants with each passing year. This (ultimately descriptive) information adds an interesting facet because it appears that although age (and hence, by proxy, technology) does affect the number of words lost each year, it does not come close to accounting for the totality of the observed downward trend, suggesting that factors beyond technology, such as structural changes in how our society congregates and communicates, are occurring and affecting everybody.

Constant dripping wears away a stone. But in this case it is not a stone we are wearing away. When we speak less, we connect less. With every lost word we wear away our connections with others. This loss is alarming because we are already in the middle of a loneliness “epidemic” (U.S. Surgeon General, 2023), in which social isolation and a loss of connection to others have become a problem of societal dimensions. By bleeding words, we bleed everyday social connection, possibly daily well-being. There are clear links between loneliness and poor health outcomes (Park et al., 2020), and there is equally clear evidence for a relationship between daily conversation quantity and well-being (Merolla & Hall, 2025).
It is imperative that we apply our best science to understand these slow, societal-level changes affecting our lives—especially because speaking 300 additional daily words may offer each individual person a trivial way to counter their personal level of isolation and thereby affect our ongoing epidemic of loneliness. For example, 300 words a day could be a brief hallway conversation with a neighbor, a colleague, or even a stranger. A joke told to loved ones. A more detailed response to the question: “How was your day?” To avoid sliding into silence, we must act now to gain back some of those 300 words we, just last year, still spoke to our fellow humans.
Acknowledgments
We thank Dan Lyons for the encouragement to write this piece and James W. Pennebaker and Steve W. Cole for their feedback on an earlier version of it. We also thank Ethan Kutlu and Gillian Sandstrom for their insightful reviews.
ORCID iDs
Footnote
Declaration of Conflicting Interests The authors declared that there were no conflicts of interest with respect to the authorship or the publication of this article.
Open practices
All data have been made publicly available via OSF and can be accessed at https://osf.io/wrtcz.
Transparency
Action Editor: Arturo E. Hernandez
Editor: Arturo E. Hernandez
References
Merolla A. J., Hall J. A. (2025). The social biome: How everyday communication connects and shapes us. Yale University Press.
Park C., Majeed A., Gill H., Tamura J., Ho R. C., Mansur R. B., Nasri F., Lee Y., Rosenblat J. D., Wong E., McIntyre R. S. (2020). The effect of loneliness on distinct health outcomes: A comprehensive review and meta-analysis. Psychiatry Research, 294, Article 113514. https://doi.org/10.1016/j.psychres.2020.113514
Tidwell C. A., Danvers A. F., Pfeifer V. A., Abel D. B., Alisic E., Beer A., Bierstetel S. J., Bollich-Ziegler K. L., Bruni M., Calabrese W. R., Chiarello C., Demiray B., Dimidjian S., Fingerman K. L., Haas M., Kaplan D. M., Kim Y. K., Knezevic G., Lazarevic L. B., . . . Mehl M. R. (2025). Are women really (not) more talkative than men? A registered report of binary gender similarities/differences in daily word use. Journal of Personality and Social Psychology, 128(2), 367–391. https://doi.org/10.1037/pspp0000534