Your Nostalgia Is Lying to You
The benefits of nostalgia are well measured. So is the inaccuracy of the memories producing them. Nobody has run the two literatures against each other
Answer to the question
Does nostalgia give you an accurate picture of the past?
No, and its benefits may depend on that. Nostalgia reliably raises self-esteem, optimism, meaning and social connectedness, yet Mitchell's 1997 study found recollection of real events rosier than the experience had been. The untested but plausible reading is that the benefit is produced by the distortion, so correcting the memory may remove the comfort.
Key takeaways
- Johannes Hofer coined nostalgia in a 1688 Basel dissertation, from nostos, homecoming, and algos, ache. He was naming a disease of Swiss mercenaries whose symptoms included weeping, refusal to eat and suicidal ideation.
- Frankenbach, Wildschut, Juhl and Sedikides pooled 116 studies of nostalgia's functions and 39 of its affective consequences in 2021, across 3,556 participants, and found robust effects on self-esteem, optimism, meaning in life, self-continuity and social connectedness.
- Leunissen and colleagues combined 41 nostalgia experiments with 4,659 participants. Inductions raised positive and ambivalent affect and did not significantly move negative affect.
- Mitchell, Thompson, Peterson and Cronk tracked people through a European holiday, a Thanksgiving break and a three-week cycling trip in 1997, rating before, during and after. Recollection was reliably rosier than the experience had been while it was happening.
- The plausible reading is not that the benefit survives the distortion. It is that the benefit is produced by it.
Is nostalgia good for you? That is the wrong question, and the research has been answering it in the affirmative for twenty years.
Wrong because it treats nostalgia as one thing. What the evidence describes is two: a feeling that reliably does something useful, measured very well, and a memory that reliably is not accurate, measured very well somewhere else by people who do not cite the first group.
Hofer
Nostalgia entered the record as a diagnosis. Johannes Hofer, a Swiss medical student, coined the word in a 1688 Basel dissertation by joining nostos, homecoming, to algos, ache. He was describing mercenaries in foreign armies who became genuinely ill with longing for home, and his case notes include emotional collapse, refusal of food and suicidal ideation. Treatment was discharge, or the Alps.
It stayed a disease for the better part of two centuries, which is worth holding onto when the current framing feels self-evident.
Sedikides and Wildschut
Constantine Sedikides and Tim Wildschut at Southampton have spent two decades reversing that, and the reversal is well evidenced rather than merely fashionable.
Julius Frankenbach, working with Wildschut, Jacob Juhl and Sedikides, pooled 116 studies of nostalgia's functions and 39 of its affective consequences in 2021, covering 3,556 participants. The main effects held across the board: self-esteem, optimism, meaning in life, self-continuity and social connectedness. Separately, Joost Leunissen and colleagues combined 41 experiments that induced nostalgia, with 4,659 participants between them, and found inductions raised positive and ambivalent affect while leaving negative affect unmoved.
Take the finding seriously, then. Indulge the feeling deliberately, because it works.
Follow that through and the question becomes what exactly is being indulged.
Loftus
Memory does not store and retrieve. It assembles, at the moment of asking, from fragments, and the assembly is shaped by present mood, present belief and everything learned since.
Elizabeth Loftus spent a career showing how far that assembly can be pushed by very little. It is worth being precise about the claim, because it gets stretched. Her work establishes that memories can be altered by information supplied afterwards, including something as light as a leading verb in a question. It does not show that old memories are generally wrong, and it licenses no verdict on any particular recollection. What it establishes is that retrieval is open to influence.
Nostalgic memories sit in the conditions where that matters most. They are old, so there is more assembly to do, and they carry emotional weight, so the assembly has a direction.
Mitchell
The direction has been measured directly, on real events, as they happened.
Terence Mitchell, Leigh Thompson, Erika Peterson and Randy Cronk followed people through three of them in 1997: a holiday in Europe, a Thanksgiving break and a three-week cycling trip in California. Ratings were taken in anticipation, during, and afterwards. Across all three, expectation ran more positive than the experience, and recollection ran more positive than the experience. The event itself was the low point of its own life cycle, every time.
Rubin
There is a second distortion underneath, and it selects what you feel nostalgic about rather than how you feel about it.
Autobiographical memories are not spread evenly across a life. Ask people over thirty for their most important memories and the returns cluster between roughly ages 10 and 30, a pattern David Rubin and colleagues established and which has since replicated across cultures and retrieval methods.
So the years most people are nostalgic for are the years they can most easily reach. Whether those years were better is a separate question the clustering cannot settle, and some of the specialness attaching to them is a fact about retrieval.
Put the two bodies of work together and the interesting possibility is not that the benefit survives the distortion. It is that the benefit requires it.
Self-continuity, the sense that the person you were and the person you are belong to one life, is easier to feel about an edited past than an accurate one. Social connectedness is easier to draw from a memory in which the people were warmer than they were. Meaning is easier to find in a period whose boredom has been deleted. Every function on Frankenbach's list is a function that improves as the record improves, and the record has been improving since the day it was laid down.
That is not proven. The two literatures have not been run against each other and the experiment would be awkward to design. But it follows naturally from both, and it is the reading that would explain why the effects are so consistent.
If it is right, the trade has no version where you get both halves. Correct the memory, insist on the anxiety and the tedium and the things you have since resolved, and the comfort may go with it, because the comfort was made of the editing. Leave it uncorrected and it stays comforting, and it stays unfit for the one use people most want to put it to: deciding whether to go back.
Which of those you want depends on what you need the past for, and that is a question about the next decision rather than about the last twenty years.
Sources
Hofer, J. (1688). Dissertatio Medica de Nostalgia. Basel. Coined the term from nostos and algos.
Frankenbach, J., Wildschut, T., Juhl, J. & Sedikides, C. (2021). Meta-analytic test of whether neuroticism disrupts nostalgia's benefits. 116 functions studies and 39 affect studies, N = 3,556; robust main effects on self-esteem, optimism, meaning in life, self-continuity and social connectedness.
Leunissen, J., Wildschut, T., Sedikides, C. & Routledge, C. (2021). "The Hedonic Character of Nostalgia: An Integrative Data Analysis." 41 experiments, N = 4,659; nostalgia inductions raised positive and ambivalent affect and did not significantly alter negative affect.
Mitchell, T. R., Thompson, L., Peterson, E. & Cronk, R. (1997). "Temporal Adjustments in the Evaluation of Events: The 'Rosy View'." Journal of Experimental Social Psychology. Three events; anticipation and recollection both more positive than the experience. Sample sizes not verified here and no figure from them is carried.
Rubin, D. C. and colleagues. The reminiscence bump: autobiographical memories cluster roughly between ages 10 and 30 for the important-memories method.
Loftus, E. F. Reconstructive memory and the misinformation effect.