When Henry David Thoreau’s pine desk arrived at the Morgan Library & Museum in Manhattan for a special exhibition on the New England writer’s life, curator Christine Nelson couldn’t resist opening one of the drawers and taking a giant whiff. The spicy terpenes from the wood, softened with age, produced an “eau de Walden Pond” that sparked images in Nelson’s mind of Thoreau sitting at the desk and writing in one of his many journals.
“When I put my head in the desk and smelled this old pine, I was overcome with emotion and a sense of connection to the past and this person who had such an effect on me,” Nelson says.
Studying the smells of history is something Nelson has recently become interested in. She’s part of a team to capture the historic odors in the Morgan library to get a better grasp of the room’s history—from the mundane act of a maid dusting the furniture to groundbreaking events, including when J. P. Morgan himself locked a group of banking magnates in the library to save the U.S. financial system.
A growing number of scientists and historians are on similar quests to deconstruct and recreate the odors of the past. Their insights could change how we think about the past and how scientists work to preserve it. For example, some odors are signs of decay in historical artifacts that conservators could then try to slow or reverse. And given the symbiotic relationship between olfaction and emotion, decoding what people might have smelled gives researchers a better idea of how they felt and what they thought.
“Smell is the holy grail of conservation science,” says Lorraine Gibson, a senior lecturer in pure and applied chemistry at the University of Strathclyde. It’s a deeper, more nuanced way to think about the past, she says.
Unlike sight and sound, which can be measured quantitatively, odor often exists in a world of metaphor and subjectivity. People describe a flooded basement as smelling like a wet dog, or the scent from a new variety of coffee as having notes of currant and chocolate.
The problem, says Michelle Francl, a computational chemist at Bryn Mawr College, is that what smells of currant and chocolate to one person may be a plain cup of joe to others. Part of our difficulty in describing smells is a lack of common vocabulary with which to do so. Classifying smells requires a standard connection with a single odor and a word, such that everyone who identifies a specific smell will use the same word to describe it. Though not as precise as using a spectrometer to measure an object’s color, standardizing odor vocabulary can give researchers a starting point.
Perfumers and vintners, whose livelihoods depend on capturing and describing precise scents, were the first to construct such a linguistic foundation. In the 1980s, Ann C. Noble, a sensory chemist at the University of California, Davis, developed a tool called the wine aroma wheel to help standardize wine descriptions, breaking down broad, easy-to-identify flavors like caramel and fruit into more precise adjectives, such as honey, butterscotch, or molasses. Aroma wheels for beer and coffee popped up soon thereafter, becoming a staple of the food and flavor industry.
Cecilia Bembibre, a graduate student in heritage science at University College London, wanted to build an aroma wheel for old book smells to help standardize descriptions of these scents, which have inspired perfumes and candles for literature lovers and library enthusiasts. For Bembibre, however, the goal wasn’t commercial; instead, she saw this vocabulary as a tool to improve conservation science.
“Many conservators can smell aged paper and gather information about a book just through experience, but there wasn’t a systematic program to identify, study, and archive these smells,” Bembibre says.
Bembibre and her adviser, Matija Strlič of UCL’s Institute for Sustainable Heritage, went to the library of Dean & Chapter at St. Paul’s Cathedral in London. They asked visitors to describe the smells that permeated the room to gather a baseline set of adjectives. One hundred percent of the participants described the room as “woody,” 86% as “smoky,” 71% as “earthy,” and 41% as “vanilla.” To construct an accurate historic book odor wheel, though, Bembibre and Strlič also needed to identify the individual chemical components of characteristic old-book smell.
Across the Atlantic, the Morgan’s Nelson had started working with techniques to identify odors. She had teamed up with master perfumer Carlos Benaim from the firm International Flavors & Fragrances (IFF) and historic preservation scientist Jorge Otero-Pailos of Columbia University to capture some of the history of the library through the odors found within. Benaim wanted to then recreate some of these odors in his lab at IFF, which meant the researchers needed a precise chemical analysis of the items they selected.
The historical importance and value of the objects at both libraries meant the researchers couldn’t take the artifacts into the lab or collect pieces of them to study. A similar problem had vexed perfumers in the past: Although they could collect objects from nature whose odors they wanted to characterize and carry them to the lab for analysis, if they wanted to capture the precise scent of a certain forest or of an unpicked flower, they had trouble. So instead of developing better ways to bring the natural world into the lab, chemists helped perfumers take the lab to nature.
In 1990, University of Waterloo chemist Janusz Pawliszyn developed solid-phase microextraction (SPME), a technique that could analyze the volatile organic chemicals (VOCs) emitted from any object. Researchers cover an item with a glass jar to trap the VOCs and then capture them using a long, thin metal needle coated with paraffin, a waxy adsorbent. After several hours, scientists insert the paraffin-coated needle into a gas chromatograph/mass spectrometer, which heats the paraffin, releasing the VOCs for analysis. Perfumers immediately began using SPME to study scents in their natural environments without needing to bring large equipment into the field. Benaim had used SPME for years and brought his expertise to the project at the Morgan.
Similarly, Bembibre and Strlič captured the molecules responsible for the scents of the old books in the library at St. Paul’s using a variant of SPME called headspace. The pair trapped the VOCs given off by an individual book by putting it in a clean plastic bag containing a carbon sponge to absorb the compounds. To sample a large space, such as an entire room, Bembibre simply sat out the carbon sponge to absorb the odors that wafted by. Bembibre’s colleagues at Odournet, a European sensory research company, subsequently analyzed the carbon sponges with GC/MS.
They also analyzed the captured odor molecules in what has come to be known as an olfactogram, a version of a standard gas chromatogram in which the peaks are annotated with the odor of each chemical as it comes off the column, as reported by a trained sniffer. The sniffers wear blindfolds to help them concentrate on scents that they describe as “vanilla” and “woody,” for instance. “It’s a very intense process,” Bembibre says, with each session lasting only 15 minutes to minimize fatigue.
Although the GC/MS can identify a chemical and correlate it with its characteristic odor in this way, the technology still hasn’t surpassed the sensitivity of the human nose. In Benaim’s experience, humans can detect smells at concentrations that don’t even register on a GC/MS. For some odorous molecules, especially diamines such as putrescine (tetramethylenediamine), and cadaverine (1,5-pentanediamine)—which reek of death and decay—the human nose’s sensitivity is several orders of magnitude higher than the most sensitive machines.
Using the olfactogram method, Bembibre and Strlič created their old-book odor wheel (Heritage Sci. 2017, DOI:10.1186/s40494-016-0114-1). The woody odors were thanks to the furfural in the decaying paper. d-Limonene gave the old books the sharp tang of an orange, and benzaldehyde provided rich, foodlike odors. Lactones added more fruity notes.
However pleasant it might be for book lovers, Nelson says that old-book smell is ultimately a sign of a book’s slow decay.
“There’s this wonderful romance around old books, but as a curator, that smell makes me think, ‘ooh, that book’s had a hard life,’ ” Nelson laughs.
Chemical analyses such as the ones performed by Bembibre and Strlič could provide opportunities to step in and preserve an item before its degradation passes the point of no return. The wood pulp used to make the paper in many 19th-century books contains lignin, which slowly breaks down into an array of acids. These acids further degrade the fragile paper. Analyzing the VOCs given off during this process with SPME could tip off conservators that further intervention is needed.
And these scientists aren’t limiting themselves to sniffing books. Old film negatives made of cellulose, vinyl records, and even early model spacesuits made by the National Aeronautics & Space Administration all degrade and decay over time. Strlič has been using similar odor analysis methods to characterize these processes to identify better ways to store valuable artifacts.
“The VOCs from historic plastics and paper follow similar degradation pathways of organic molecules. Conservation scientists have always been looking for less-destructive ways to sample an object to check its status, and what’s less destructive than simply smelling it?” Strlič says.
Besides aiding with conservation, this chemical analysis of odors will help researchers understand the building blocks of important historical smells, possibly allowing the scientists to recreate those scents, which have largely been lost to time.
Benaim and his team at IFF have begun trying to synthesize odors captured at the Morgan library. To do this, Benaim will have to combine the right chemicals in the right proportions to create a single, recognizable smell. For example, when German scientists analyzed the aroma molecules from roasted cacao beans, the key ingredient in chocolate, they found a combination of 25 key chemicals, including 2- and 3-methylbutanoic acids, which produce a rancid, sweaty odor, and dimethyl trisulfide, which smells like cooked cabbage. Too much of either of these, and the sweet, chocolate aroma suddenly resembles an old gym bag or overcooked sauerkraut. Even smells that the most proletarian palates can distinguish, such as coffee and chocolate, may differ not so much in their chemical building blocks but rather in the proportions of those odor building blocks.
Besides the technical difficulties of assembling a large number of molecules in the proper proportions, historian Mark Smith at the University of South Carolina cautions that interpreting these odors may be more difficult still. In previous eras, the omnipresent stench of unwashed bodies, manure, rotting fish, and wood smoke formed a nearly unnoticed olfactory backdrop that would likely overwhelm modern noses. Even if scientists recreate this odor landscape down to the last molecule, a 21st-century American will have a different experience with each inhale than a ninth-century Viking or a 17th-century Parisian. What we smell today will have an entirely different meaning to what they smelled back then.
“You have to ask whether your act of smelling something is the same as it was for them,” Smith says. The effort of interpreting historic smells “tells us more about us than it does the past.”
One example of our changing odor landscapes is the famous potpourri found at Knole, a 600-year-old manor house in Kent, southeast of London. Recipes helped historians create an accurate reproduction that is now for sale in the house’s gift shop. But Bembibre says that the potpourri, whose recipe dates to the 1750s, doesn’t always appeal to modern noses. The combination of dried flowers from the Knole garden, including lavender, bay leaves, and geraniums, and spices like mace and cinnamon, are foreign to some modern visitors, who aren’t accustomed to this combination of scents.
Despite the seeming banality of a bowl of potpourri or a pine writing desk, Bembibre and Nelson say that scents like this are worth preserving because they can illuminate the past in ways that pictures and documents can’t. It’s why Nelson felt compelled to sniff the drawers in Thoreau’s desk.
“Preservation is about memory, and what’s more profound than smell?” she says.
Carrie Arnold is a freelance science writer based in Richmond, Va.
CORRECTION: This story was updated on Nov. 29, 2017, to correct the molecular “category” of putrescine and cadaverine. These molecules are diamines, not thiols.