Present Past: Aroma of A Memory

It's often difficult to know just what provokes a memory.  A smell can, but it's complicated, and even involves something that looks like a paradox.

In his chapter on the subject in Why Life Speeds Up As You Get Older, Dowe Draaisma notices a complication, which he elegantly frames with an observation about the famous petite madeleines scene in Proust.  Every psychological treatise on smell and memory inevitably refers to it, and like good English majors, they repeat what they've heard about it (probably from other treatises on smell and memory) and they generally get it wrong.

"The depiction is often at umpteenth hand, three lines long at most and whittled away until almost unrecognizable: the narrator drinks a cup of tea, dunks a piece of cake in it, and suddenly the smell takes him back to his youth in Combray."

The problem is the actual scene as Proust writes it covers four pages and goes through a long process between the smell and the memory.  The smell affects him but he doesn't know why.  Only with effort does he then associate it with the general memory of his childhood.  It's a key to many memories of that time and place.

A smell can ignite a memory pretty much right away, and it can be powerful.  That's happened to me.  But it also--and probably more often--just causes a slight change in mood, the research suggests.  That may lead to a memory, but only with some work at concentrating, sorting, rooting around--as Proust did.

Smell has several unique features compared to other human senses.  Of course, it's pretty weak in humans, compared to other animals and to other senses.  Taste is even more specific, and much of what we think of as taste is actually (or mostly) smell.

And smell has a unique pathway to the brain, a direct route to centers of emotion and memory.  But shouldn't that make remembering smells more likely?  Not necessarily, for autobiographical memory is strongly linked to speech, and since smells bypass speech centers, we may have no access to those memories.

But when an aroma does evokes a memory it can be very powerful.  Accounts by writers suggest these would more likely be smells that were a fairly regular part of life in the past but are not frequent in the present, like the smell of hay that immediately takes a city dweller back to childhood on the farm.

Memories--not usually of specific events but of a period or place--sparked by smells are usually positive, Dr. D. writes, occasionally negative but almost never neutral.  It's that direct line to the limbic system, where emotions reside.  And the sight of hay or sawdust or wool tweed is not enough to trigger the sensation.  It has to be the smell.

Lab experiments trying to verify smells evoking memories have generally failed, because you can't get enough smells into the laboratory, and because it may take time before the smell and memory are connected.  Another reason they fail apparently is because experimenters use the same undergraduate age test subjects as many if not most of these studies do, and then they announce findings as if they necessarily apply to all ages, cultures, etc. in all conditions.

But one lab study that tested subjects of various ages found that people over 70 were much more likely to have memories elicited by actual smells than the names of smells.  The apparent paradox is that the sense of smell declines much more with age than other senses.  70 year olds typically have a small fraction of the ability to smell as they did as children.

But that doesn't undercut the power of particular smells; it may enhance it.  For older subjects, smells were associated mostly with memories formed when they were age 6 or younger.  It's that direct connection again.  And this effect is enhanced by another unique property of the smell sense.

Our active memory can hold only so much--so many thoughts, associations, sights, sounds, etc.  So newer sights tend to block the memory of older ones, which is true of the other senses--except smell.  New smells don't block old smells.  So another reason that, if a smell does provoke a memory, it's a powerful connection.

Dr. D. doesn't deal with a phenomenon I've noticed: certain smells can evoke a powerful sense of "being there" in the memory, and it can happen each time for that smell.  But if the smell isn't repeated, I forget the association.  I can remember only a few recent smells that have taken me back, but I know there have been others.  I just don't remember them.

Obviously an aroma you haven't smelled in awhile is more likely to bring back an emotion-laded memory.  But I've noticed that it's more than a simple correspondence, especially for smells more regularly encountered.  For example, concrete blocks and bricks.  Only sometimes in the presence of concrete and bricks do I catch the aroma that takes me back to houses under construction in my neighborhood when I was a kid (positive.) Or sites where I later worked as a summer job, mostly picking up and stacking in the hot sun (negative.)

Dr. D. also doesn't deal with self-reported associations.  For that I go back to something I clipped out of Harper's Magazine (I believe) several decades ago.  I can't find the actual clipping, but I wrote down the findings.  It was a survey done in a Chicago shopping mall.  A social scientist asked people "What odors cause you to be nostalgic?"  What was most interesting about the findings--and from a certain point of view, sad-- was how they broke down in terms of when people who had them were born.

Among the smells people born in the 1960s and 1970s cited were: Play-Doh, Chlorine, crayons, Downy fabric softener, tuna casseole, tacos, SweeTarts, Coca Puffs, Scented Magic Markers, Windex, hairspray, disinfectant, motor oil, airplane fuel, plastic and smoke.

Among the smells those born in the 1920s, 30s and 40s cited were: lilies, manure, violets, ocean air, pine, hot chocolate, Cracker Jack, baking bread, blueberries, honeysuckle, burning leaves, clover, hay, meadows, tweed, meat balls, cut grass, soap, fresh air.