Temple Grandin on problem-solving, research and life under coronavirus

Pictured above: Dr. Temple Grandin with Scatter, a 25-year-old Appaloosa gelding, on the ranch of her longtime research partner, Mark Deesing, in Fort Colllins, Colorado (Photo © Callen Liverance(opens in new tab/window))

Dr. Temple Grandin pictures her immune system as a military base with soldiers who defend the base and fight off invaders. “Unfortunately, some of the soldiers are stupid,” she says, “and when they get overwhelmed, they go crazy and start to burn down the barracks and trash the base.”

She goes on to explain how some drugs are designed to stop the invaders and others to put out the fires.

“I am extreme visual thinker,” she says, “and all my thoughts are in photo-realistic pictures. To think about a concept, I have to convert it to a pictorial analogy.”

Analogies are among the techniques she uses to make sense of abstract concepts and solve complex problems.

Using visual analogies is something that comes easily to her as an autistic person who initially struggled with language and didn’t speak until she was almost 4. Over the years, Temple would learn to compensate by using her strengths, with extraordinary results.

Her ability to “think in pictures” instead of words helped her develop an uncanny understanding of animal behavior. She went on to become an animal scientist with a world-renowned business designing more humane livestock handling systems.

“This kind of thinking helped me to be successful … because I could think about the animals without using words,” she says. “And it’s also helped me to be good at practical problem solving because I can visualize a solution.”

Temple speaks emphatically into her iPhone, as though she wants to make sure her ideas are clear to me 1,700 miles away in New York. Actually, it’s the same tone she uses in her classes at Colorado State University, where she is Professor of Animal Science, or at the dinner table, or in her autism lectures, which she gives in packed auditoriums around the world.

As an autism advocate and author, she gave her last lecture earlier this month before her remaining speaking engagements were cancelled amid the looming pandemic. On a near-empty passenger plane from New Jersey, she hunched over her iPhone studying the latest news and research about COVID-19 treatments.

It’s become a fixation for her, combing through articles in scientific journals and pharma trade publications, scrutinizing the methods, and combining the findings with what she observes in the media and the world around her.

She wants patients to get the right medications at the right time, before it’s too late. She’s concerned about her 93-year-old mother, who lives on her own in Manhattan. “And also I’d like to not die myself,” she adds, mentioning that she’s 73.

Temple’s quick to point out that she’s not a medical professional, just a darned good problem solver. And in the throes of an unprecedented pandemic, problem solving is at the heart of what researchers are doing in labs around the world.

Also, like so many researchers whose expertise is elsewhere, she desperately wants to contribute.

Her world, like ours, has been transformed by coronavirus. So she’s staying put for now, driving between her home in Fort Collins, Colorado, and the nearby ranch of her longtime research partner and co-author Mark Deesing. While sheltering in place, Mark does Temple’s grocery shopping, and Temple helps take care of his horses. Meanwhile, they are preparing the third edition of their book Genetics and the Behavior of Domestic Animals, published by Elsevier.

On a break, she reflects on her problem-solving techniques, which she hopes will be useful for scientists whatever they’re working on.

1. Think from the bottom up.

Temple remembers her earliest attempt to make sense of the animal world:

When I was a kid, I would sort cats, dogs and horses into three categories: small, medium and large. Okay, now that way of categorizing got ruined when a dachshund came into the neighborhood the same size as cats. I remember sitting on the lawn of the next-door neighbor's house studying Rosie the dachshund and trying figure out why she was a dog.

So she set out to find features Rosie shared with other dogs, like barking: “In other words, she makes the sound of a dog, she has the smell of a dog and the nose shape of a dog,” Temples says.

These are things I was figuring out at 6 years old. This is the very beginning of bottom-up thinking.

Bottom-up thinking starts with collecting data: it can be facts and statistics, or it can be observations from our everyday lives. In Temple’s words, “You take a bunch of data about something you don't know about, and you (find patterns to) sort them into categories.”

Some AI methods use a similar approach, Temple explains. She mentions a 2017 study by Stanford researchers(opens in new tab/window) who trained a deep-learning algorithm to diagnose skin cancer by classifying skin lesions as benign or malignant. The researchers used smartphone photos to make the study relevant to the many dermatologists who now use their devices for this purpose. As a result, the categorization process was more difficult because of differences in lighting, angles and quality. “We overcome this challenge by using a data-driven approach — 1.41 million pre-training and training images,” the authors wrote.

As with the computer algorithm, constantly adjusting its assessment of what malignant looks like, bottom-up thinking requires – and encourages – a flexible mind. “As you get new data, you're much more willing to change the hypothesis,” Temple says.

For a scientific study, the bottom-up process involves collecting data before forming the hypothesis and designing the experiment. This is how Temple designed her studies for livestock handling systems. Some of her data consisted of her own observations of what “spooked” animals – like a waving flag or a reflection in a puddle or a “black hole,” where it’s sunny outside, and livestock don’t want to go into a dark building – and also what made them comfortable. Often, these were the same things she experienced as an autistic person: she was easily shocked by sudden movements and high-pitched noises and strangely comforted by the “squeeze machine(opens in new tab/window)” she invented as a teenager.

Temple designed her livestock systems with these principles in mind, traveling to sites that use them to show people how to manage their cattle more efficiently and humanely. Despite initial skepticism in the industry, these systems resulted in cattle moving through systems much more easily than the older equipment, avoiding injuries. Now, the equipment she designed is used in most of the large beef processing plants in North America.

Bottom-up thinking is the opposite of top-down thinking, where the hypothesis is formed and then data is collected that may or may not support it. The top-down approach is a cornerstone of scientific research and has many benefits. It’s how controlled clinical trials of new drugs are structured, for example. These studies rely on strict protocols and an unbiased approach by researchers.

There are pitfalls, however. For instances, researchers may choose to withhold “negative” results that do not support the hypothesis – either because they don’t think they’re important or because they don’t support the hypothesis they had hoped to prove. This reality has led to a movement to publish negative results as relevant information for researchers and the public.

Another important factor today is that the top-down method takes much longer. That’s why, with researchers racing against the clock to find the best treatment for coronavirus, they’re taking a bottom-up approach, Temple explains: they’re looking into repurposing existing drugs with known risks and benefits rather than developing brand new ones.

There's a practical side to Temple's perspective that comes from her years in business, where cost, supply chain and ease of manufacturing determine whether a scientific vision can be turned into a viable solution.

"In my equipment design work, I have always told myself to think simple," she says. "These drugs should be cheap and easy to manufacture, as opposed to requiring some fancy dancy process that needs a trillion-dollar lab to make it."

2. Develop your visual thinking skills.

The bottom-up approach to research benefits greatly from keen observation skills.

Recently, Temple was flabbergasted when her research partner returned from the local pharmacy and told her about his experience.

As Mark explained, the store had taken all kinds of measures to germ-proof their premises, sanitizing the aisles and installing plexiglass screens at the checkout counters. But when he pulled up to the drive-through window to pick up his prescription, he would have had to pick up an old-fashioned telephone receiver that everyone else had touched and put it to his mouth to place his order. He ended up parking and going inside.

“That’s positively crazy,” Temple says. Years ago, she would have called the people who make this oversight “stupid,” but now she knows better: “Some of the things I was calling people stupid for are visual thinking mistakes,” she explains, “because I didn't know at the time that other people were not visual thinkers.”

The power of observation is behind many scientific breakthroughs. As an example, Temple mentions the famous Broad Street cholera outbreak of 1854(opens in new tab/window). With frequent outbreaks around London claiming thousands of lives, scientists debated two theories on the cause of cholera: germ theory or miasma theory(opens in new tab/window), which claimed that diseases like cholera and plague were caused by noxious fumes or “bad air” from rotting organic matter.

From her years of teaching, Temple says she realizes that many people are not visual thinkers, or they use their stronger senses at the expense of their visual ability. She explains this theory eloquently in her 2010 Ted Talk: The world needs all kinds of minds(opens in new tab/window).

If her speaking seems too polished and fluent for one who once struggled with language, Temple shares her secret. “I talk through pictures” – pictures that flash through her mind “like Google Images.”

To cultivate our visual ability, Temple recommends practicing by visualizing spaces we already know – like our desk at work or our local grocery store. Picture as many details as you can before comparing your view to what is actually there. She shares an example from her own life:

At Whole Foods, they have these dark brown wooden square boxes they set the produce on. Okay, I'm seeing that right now. And I'm seeing where they put the little (shopping) baskets right by the sliding door. But the thing I'm really seeing there is that dreadful credit card machine with that pen that everybody touches. That is the germ center. … 

“And now I'm seeing another thing,” Temple continues:

There's a lady next door that works for a delivery company. And I saw her last night, and she had her shirt on for DHL, right? We're outside, 20 feet (from each other), so we're safe. And we were talking about how she keeps safe.

Temple deepens her voice to a dramatic stage whisper to share her neighbor’s secret:

Nobody touches my scanner. Nobody. I sign for them all with an X.

“Everyone’s worried about getting cooties from cardboard boxes,” Temple adds, “but it’s those touch screens and pens they should really worry about. .. You get rid of the shared pen, it's like taking the handle off that pump."

3. Don’t forget to check the methods – and make yours reproducible.

While her life has been fraught with challenges, 15 years ago, Temple encountered a problem that was especially daunting: she started to lose her hearing in one ear. She was diagnosed with an autoimmune disease called Meniere’s disease(opens in new tab/window), an inner-ear disorder that can lead to hearing loss.

She left the doctor’s office with a prescription for a diuretic, which did not stop the hearing loss – and a mental picture of “the huge hearing aid store attached to his office.” After finding another doctor, she started combing the scientific literature for treatments for a related illness, sudden sensorineural deafness(opens in new tab/window). What she found might have confounded some. In some studies, a certain drug worked, in others it failed. “I’m going through 20 abstracts: It worked, it did not work, it worked, it did not work, back and forth.”

While some people tend to blame the scientists – or science itself – when results don’t agree, that’s not Temple’s way: “I’m not going to call anyone’s paper crap; I’m going to look at the data and figure out why the results are different.”

For her, the answer is often in the methods section:

One basic thing that I have found is that when you get studies that don't agree – I don't care what the field is – you look at the methods. Then you'll find things like they used a different breed of pig, for example, or they were housed differently. With autism stuff, the way they sampled the population was different. Maybe they had an older population or a younger population. That can explain why some results came out opposite. I’ve seen that over and over again.

Methods can be an important source of data in themselves. Two years ago, Elsevier Labs data scientists conducted a meta-analysis(opens in new tab/window) that led to a finding with important implications for cancer research. Previous studies had shown that the temperature of lab mice affects their stress levels, and that mice show resistance to chemotherapy in lower temperatures. To see if scientists were changing the way they housed mice in light of these findings, they examined methods data from about 8 million research papers with cancer mice models and chemotherapy drugs. They found that sub-optimal mice housing conditions were still being used in cancer experiments.

“This is probably responsible for significantly slowing down the science of cancer biology,” says co-lead author Dr. Helena Deus.

“In many studies,” she adds, “this (temperature) information was not even included in the paper.”

Methods without relevant data are a source of frustration for many researchers. As a reviewer of animal science papers, Temple says some researchers leave out important details, like the breed and age of the animals, how they were housed and what they were fed. Her advice to researchers:

You have to include enough details so that someone can replicate the experiment.

In researching the treatment for her hearing loss, she found that in some studies, the drug in question was given at an earlier stage of the illness, in others at a later stage.

In her case, it was too late to repair the damage, which caused her to lose about 50 percent of her hearing in the affected ear, but the medication she ended up taking prevented further damage.

4. Use visual analogies.

Temple sums up her hearing loss story with a simple analogy:

The lightbulb moment was that it’s the timing you give the drug: It puts out the fire, it doesn’t repair the damage.

For Temple, who has always struggled to grasp abstract concepts, analogies give form to ideas and help her visualize them. While some autistic people are too literal to understand analogies and metaphors, that’s not the case for her. For example, she loves the saying “strike when the iron is hot(opens in new tab/window)” – a metaphor from blacksmithing that reminds us to take action while the opportunity is still there.

Another metaphor she uses involves doors. For her, the idea of overcoming challenges to progress in life is too abstract. So she envisions herself walking through a door to enter each new stage of her life and career – a metaphor vividly conveyed in the HBO film about her early life(opens in new tab/window).

In talking about medicine, she uses her analogy of the military base to emphasize that different drugs should be taken at different stages depending on their purpose:

There are drugs that put out the fire of the inflammation, but they don’t repair the damage. If an anti-inflammatory drug is going to work, it has to be given before the military base is completely wrecked. Anti-inflammatory drugs force your crazy soldiers to stop and stand down.

Then she offers an even simpler analogy:

If I have a wastebasket fire, the drug puts it out. If the house is on fire, I’m screwed.