What the Nose Knows: The Science of Scent in Everyday Life (11 page)

S
O CLOSELY IS
sniffing tied to odor perception that people routinely sniff when they are asked to imagine a smell. Without prompting, they take larger sniffs when imagining pleasant odors and smaller ones when imagining malodors. During visual imagery the eyes explore an imagined scene using the same scan paths made when viewing the actual visual scene. Preventing eye movements during visual imagery—by having people stare at a stationary target—reduces the quality of the image. Sobel found that, similarly, imagined odors were more vivid when people could sniff than when they were wearing nose clips and unable to sniff. Actually sniffing increased the unpleasantness of an imagined bad smell (urine) and increased the pleasantness of a good one (flowers). Sniffing at an imaginary odor isn’t an absentminded habit—it’s a behavior that improves the mental image we are trying to create. Sobel’s claim that “the sniff is part of the percept” would have outraged Charles Elsberg, but it sounds reasonable to most neuroscientists today.

We have in fact done a complete about-face since Elsberg’s attempt to measure smell without sniffing. Because smelling
is
sniffing, we can now test odor perception by measuring sniffing alone. We can take advantage of the fact that people naturally and unconsciously take smaller sniffs when an odor is present: the stronger the odor, the smaller the sniff. People with no sense of smell fail to adjust; they keep inhaling as if the air were unscented. A new smell test, developed by University of Cincinnati psychologists Bob Frank and Bob Gesteland, is simplicity itself. The patient wears a pair of standard-issue medical nose tubes connected to an electronic console, and sniffs at half a dozen cylinders in a row. That’s it—test over. No need to identify smells by name, no multiple-choice questions, no rating scales, no fancy odor generators. Here’s how it works: Each cylinder is the size of a can of beans and may or may not contain a slightly unpleasant odor (in pilot testing, Frank and Gesteland used methylthiobutryate, which has the character of feces, putridity, decay). The test console records airflow into the patient’s nose and computes the size of each sniff. It compares sniffs made when the patient was smelling scented cylinders with those made to an empty cylinder. If the two types of sniff are of similar size, the patient almost certainly has an impaired sense of smell.

Remedial Sniffing

We have glasses to help those with defective vision, hearing aids for the partly deaf, and who now will produce an artificial device to improve the smelling ability of people with subnormal noses?

—Popular Science Monthly
, 1931

If perception and sniffing are inseparable, what happens to people who can’t sniff? The most extreme case of nonsniffing is the person with a total laryngectomy, or removal of the voice box (larynx), a procedure that disconnects the upper and lower respiratory airways. After laryngectomy, a person breathes through a hole in his throat, rather than through the mouth or nose, so he is unable to sniff or even activate his vocal cords to speak. Adding to their misery, about 85 percent of these patients are smell-impaired. Fortunately, some can be helped by a simple physical maneuver that resembles a polite yawn, or in other words, yawning with the mouth closed. This pseudo-sniff technique pulls air through the nose (though not the lungs) and allows about 50 percent of patients to score in the normal range on a smell test. A device called a tracheostomy valve, which directs exhaled air upward past the vocal cords and into the back of the nasal passages, restores speech function and also improves odor perception.

Impaired sniffing also occurs in Parkinson’s disease and contributes to the smell loss found in these patients. Because the disease affects motor movement, the sniffs of a Parkinson’s patient are weak and small. The worse their sniffing, the worse their performance on olfactory tests. The patients with the worst deficits can improve their test scores by simply taking bigger sniffs. While part of the problem lies in the physical action of the sniff, Parkinson’s patients often develop cognitive impairment, which registers on smell tests; in fact, smell deficits are an early symptom of the disease.

A 1996 U.S. patent describes a device to help the sniff-impaired. It resembles a double-ended turkey baster, with the bulb in the middle equipped with one-way valves. The user positions one end of the device over, say, a bowl of chili, then squeezes and releases the bulb, and it fills with air. Now the user inserts the other end in his nostril and squeezes again, forcing a bulb full of chili-scented air up his nose. The device is sort of an Elsberg self-blaster, a nose trumpet for the hard of smelling.

Boosting nasal airflow even improves odor perception in normal people. The Breathe Right nasal dilator was first marketed in 1993 to help reduce snoring by increasing nasal airflow, but got attention as an athletic aid the following year when Herschel Walker of the Philadelphia Eagles wore one for the first time in an NFL game—he had a cold. When Jerry Rice of the San Francisco 49ers followed suit, the Breathe Right gained locker-room cred, and commercial success followed in drugstores across the country. The dilator is placed on the bridge of the nose just above the fleshy portion of the nostrils, where it exerts a springlike action that prevents the sides of the nasal vestibule from collapsing inward during an indrawn breath. (The nasal vestibule is the space behind the opening of the nostril; it’s the finger-pickable part of the external nose.) Testing shows that wearing a dilator makes odors smell stronger, improves odor identification ability, and helps the wearer detect an odor at significantly lower concentrations. These benefits are the result of more air getting up into the nose. The nasal dilator increases the intensity of food aromas in the mouth but, weirdly, decreases the pleasantness.

T
HE ACT OF SNIFFING,
overlooked by many scientists and politely ignored by well-mannered people, is critical to how we generate a mental image of the smellscape. The rapid sampling of odor-laden air is managed by a precisely timed interplay of sensory and motor function. In many instances, sniff improvement results in smell improvement. Seventy years after Charles Elsberg set out to suppress the sniff, we have finally begun to appreciate its value.

Even as it makes midsniff adjustments to the smell stream entering the nose, the brain is actively fine-tuning the mental impression it creates from an odor through a process called adaptation. Everyone is familiar with visual adaptation: after being in bright sunlight, it takes a minute or two for your eyes to adjust as you enter a darkened room. The reverse happens when you leave a movie theater in midday: the sunlight is unbearably bright at first, but gradually you adjust. Olfactory adaptation works on a similar principle: a new odor smells strong when we first experience it, but the longer we’re exposed to it, the more it fades into the background. In the extreme, the smell may be undetectable for a while.

It’s easy to overstate the practical importance of this phenomenon. Adaptation is a temporary change; it doesn’t permanently erase the ability to smell. Fragrances are not written in disappearing ink: if women stopped smelling an eighty-five-dollar perfume within a few days of buying it, the fragrance industry would have collapsed long ago. The extent of adaptation depends on the nature of the smelling being done. Perfumers I know insist they can only smell half a dozen fragrances before they notice a dulling of perception. For these professionals, olfactory fatigue is a real obstacle. They sample trial perfumes from blotters, five-inch strips of filter paper dipped in the liquid. The professional takes a quick sniff or two and moves the blotter away, ever conscious of overdoing it.

In contrast, an amateur sniffer holds the blotter in front of his nose and inhales continuously, a sure-fire way to dull the nose. Even one minute of such deep breathing makes an odor immediately harder to detect. When I run a consumer smell test, I let the panelists sniff at their own natural pace. I’ve found they can easily assess a couple of dozen scents without a noticeable decline in performance. That’s because they are sampling a variety of scents and doing so to make a quick thumbs-up or thumbs-down opinion—the typical objective of consumer and market research. This poses much less risk of adaptation than does the perfumer’s repeated study of minor differences between related samples. The average person making rapid-fire judgments does not need to worry about the smellscape fading from view.

T
HE LONGER YOU
are exposed to an odor, the more you adapt to it. Step into a garlic factory and the reek will overwhelm you. A few minutes later its intensity fades, and after an hour you might not be able to smell garlic at all, no matter how hard you try. Work there a few months and this adjustment will happen almost as soon as you step in the door. That was how I once became oblivious to
Safari.
Early in my career, the company I worked for was developing the perfume for Ralph Lauren. As we tweaked the formula, ran stability tests, corrected the color, and did the million other chores needed to ensure a successful launch, the entire building was steeped in
Safari.
A few weeks into the job, none of us noticed it.

After a long vacation, I opened my closet to grab a suit for work, and got an overpowering faceful of
Safari.
The sensory truce between my nose and my workplace had fallen apart in less than two weeks. Similarly, long-term adaptation is what keeps plumbers and pig farmers from going insane.

Adaptation is a two-way street: when the odor source is removed, the nose gradually regains its sensitivity. This time-course of recovery is almost the mirror image of adaptation. Step outside after your visit to the garlic factory, and the recovery begins. If you were inside for just a few minutes, recovery will take a matter of minutes. If you were there for hours, it will be hours before full response returns. Odor strength is another factor in adaptation. The stronger the smell, the more you adapt. Ten minutes on the processing floor of the garlic factory will cause more adaptation than ten minutes talking to someone with garlic breath.

Adaptation is also odor-specific. If you work in a garlic factory, your nose will selectively tune out garlic, but your sensitivity to roses, sour milk, beer nuts, and other un-garlic-like smells will be unaffected. The narrowness of adaptation is sometimes exploited by perfumers when they try to match one fragrance to another. A perfumer will use saturation sniffing as the final step in comparing the target and the make. He sniffs the sample to the point of total adaptation, then smells the target; with his brain filtering out any sign of the original, any remaining minor differences will stand out.

Adaptation is a useful feature of any sensory system; it preserves our ability to detect small differences between stimuli against enormous variation in overall intensity. Just as auditory adaptation lets us have a whispered conversation but also talk in the middle of a rock concert, olfactory adaptation constantly recalibrates our noses to background conditions. It also selectively tunes new smells into the background, freeing our attention for the next new scent that may be creeping our way.

The Spin Doctors

In a lecture hall at the University of Wyoming in 1899, a chemistry professor named Edwin E. Slosson played a prank on one of his classes. He explained that he wanted to demonstrate the diffusion of odor through the air. He poured some liquid from a bottle onto a wad of cotton, making a show of keeping it away from his nose. He started a stopwatch and told the students to raise a hand as soon as they smelled something. Here’s what he reports happened:

While awaiting results I explained that I was quite sure that no one in the audience had ever smelled the chemical compound which I poured out, and expressed the hope that, while they might find the odor strong and peculiar, it would not be too disagreeable to anyone. In fifteen seconds most of those in the front row had raised their hands, and in forty seconds the “odor” had spread to the back of the hall, keeping a pretty regular “wave front” as it passed on. About three-fourths of the audience claimed to perceive the smell, the obstinate minority including more men than the average of the whole. More would probably have succumbed to the suggestion, but at the end of a minute I was obliged to stop the experiment, for some on the front seats were being unpleasantly affected and were about to leave the room.

Slosson’s experiment vividly demonstrated the potency of olfactory suggestion, for he was holding a cotton ball soaked in nothing but water.

The sensory expert Michael O’Mahony revisited the phenomenon in the late 1970s. During a British television documentary on taste and smell, he showed viewers an electronic device that he claimed could capture and broadcast odors using “Raman Spectroscopy.” The machine played a ten-second audio tone that viewers were told would evoke a “pleasant country smell.” They were encouraged to call in or write and describe what they smelled. Many did. They reported smelling new-mown hay, freshly cut grass, lavender, honeysuckle, and so on. O’Mahony repeated the trick on a BBC radio show using a supposedly inaudible “ultra high frequency tone”—actually no sound at all. Some listeners reported smell sensations when it was played.

While amusing, these stunts by Slosson and O’Mahony raise serious questions for scientists conducting smell studies, because they show that just expecting a smell can trigger an odor perception. Thus a purely psychological expectation might have the same consequences as a real smell. For researchers the question becomes, How can we be sure the results of an odor experiment are really due to the smell and not to expectations about the smell? What is needed is an olfactory placebo: a test condition in which people are led to believe an odor is present when in fact it is not. To truly have an effect, an odor must outperform the placebo. This was the reasoning behind a study I did with Susan Knasko, a postdoctoral fellow of mine at the Monell Center, and the late John Sabini, a psychology professor at the University of Pennsylvania. We sprayed water mist in the air and told people it had a smell. The test room was actually scent-free and remained so. People who were told the smell was unpleasant later rated the room as smelling bad. When told the smell was pleasant, they liked the smell of the room. A supposedly “neutral smell” produced intermediate results. Interestingly, physical symptoms such as headache and itchy skin were also affected by the “good smell” and “bad smell.” Our study was the first to confirm in the laboratory that the power of suggestion, by itself, could produce odorlike effects.