About the SA Blog Network

Guest Blog

Guest Blog

Commentary invited by editors of Scientific American
Guest Blog HomeAboutContact

Lessons from Sherlock Holmes: Don’t Judge a Man by His Face

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

How do we perceive someone we’ve only just met? How do we judge him, assign him to some sort of category in our mind, explain to ourselves what he is and what he is likely to be?

In “The Adventure of the Norwood Builder,” Dr. Watson demonstrates an approach that we are all too likely to follow naturally: judging too quickly from our initial impressions and failing to correct for the specific circumstances involved. After Holmes enumerates the difficulties of the case and stresses the importance of moving quickly, Watson remarks, “Surely the man’s appearance would go far with any jury?” Not so fast, says Holmes. “That is a dangerous argument, my dear Watson. You remember that terrible murderer, Bert Stevens, who wanted us to get him off in ’87? Was there ever a more mild-mannered, Sunday school young man?” Watson has to agree that it is, in fact, so. Many times, people are not what they may initially be judged to be.

When person perception breaks down

The process of person perception is a deceptively straightforward one. First, we categorize. What is the individual doing? How is he acting? How does he appear? Then, we characterize. Ok, now that I know what he’s doing or how he seems, what does that imply? Are there some underlying traits or characteristics that are likely to have given rise to my initial impression or observation? Finally, we correct: is there something that may have caused the action other than my initial assessment (in the characterization phase)? Do I need to adjust my initial impressions in either direction, augmenting some elements or discounting others?

So far, so good. Except, there’s one major problem: while the first two parts of the process are nearly automatic, the last is far less so–and often never happens at all.

According to psychologist Daniel Gilbert, we can be passive or active perceivers. As passive perceivers, we merely observe, but as active perceivers–the state we are in in most social situations–we are always multitasking, trying to navigate the complexities of social interaction at the same time as we make attributional judgments. In a series of classic studies, Gilbert set out to demonstrate that active perceivers categorize and characterize on a near subconscious level, automatically and without much thought, but then fail to implement the final step of correction–even when they have all of the information to do so–and so end up with an impression of someone that did not take into account all of the variables of the interaction. Like Watson, they remember only that a jury would like a man’s appearance; unlike Holmes, they fail to take into account those factors that might make that appearance a deceptive one – or those circumstances under which a jury would dismiss any appearance, no matter how trustworthy, as false.

In the first study, Gilbert and colleagues tested whether individuals who were cognitively “busy,” or multitasking in the way that we often do when we juggle numerous elements of a situation, would be able to correct initial impressions by making the necessary adjustment. A group of participants was asked to watch a series of seven video clips, where a woman was having a conversation with a stranger. The clips did not have sound, ostensibly to protect the privacy of those speaking, but did include subtitles at the bottom of each clip that told participants the topic of conversation. In five of the seven videos, the woman behaved in an anxious fashion, while in the other two, she remained calm.

While everyone watched the exact same videos, two elements differed: the subtitles and the task that the participants were expected to perform. In one condition, the five anxious clips were paired with anxiety-provoking topics, such as sex life, while in the other, they were paired with neutral topics like world travel. And within each of these conditions, half of the participants were told that they would be rating the woman in the video on some personality dimensions, while the other half was expected to both rate personality and be able to recall the seven topics of conversation in order.

What the researchers found came as no shock to them, but did shake up the way that person perception had always been seen. While those individuals who had only to focus on the woman adjusted for the situation–rating her as dispositionally more anxious in the neutral topic condition and as less anxious in the anxiety-inducing topic condition–those who had to recall the conversation topics completely failed to take those topics into account in their judgment of the woman’s anxiety. They had all of the information they needed to make the judgment–but they never thought to use it. So, even though they knew that the situation would make anyone anxious in theory, in practice they simply decided that the woman was a generally anxious person. What’s more, they predicted that she would continue to be anxious in future scenarios, regardless of how anxiety-provoking those scenarios were. And the better they recalled the topics of conversation, the more extremely their predictions were off. In other words, the busier their brains were, the less they adjusted after forming an initial impression.

In the second study, Gilbert addressed two remaining concerns: would the effect hold for observing verbal behavior, which is less automatic than physical, and would it hold in a more real-world applicable scenario, such as arises when we’re talking with someone and getting ready to say something ourselves in response? Here, participants listened to a man give a speech that he had been randomly assigned to write. The speech was either pro- or anti-abortion. The task: to guess at the real views of the person reading the speech, even though they may be opposed to the words themselves.  Half of the participants could just listen; the other half was told that they themselves would be asked to write and read a speech later in the study session.

What happened was analogous to the prior results: those individuals who knew they would give a speech later on were more likely to attribute the views of the speech at hand to the participant himself, believing that what he said corresponded to his actual thoughts despite knowing that the assignment had been random. They knew, and at the same time, they didn’t know. They understood in theory, but at the same time, they did not make the translation from theory to practice. It was too much cognitive work. And they had enough to think about as it was.

What this implies for our daily interactions

The news here is both good and bad. First, the obviously bad: in most situations, under most circumstances, we are active observers, and as such, more likely than not to make the error of unconsciously, automatically categorizing and characterizing, and then failing to correct that initial impression. And so, we go by appearances; we forget to be subtle; we forget how a person can easily be influenced at any given point by myriad forces, internal and external. Incidentally, this works whether or not you tend, as most westerners do, to infer stable traits over passing states, or, as many eastern cultures do, to infer states over traits: whatever direction you err in, you will fail to adjust.

But there’s good news. Sometimes, it’s good to be cognitively busy and not have time to make adjustments. It’s important to remember that there is a hierarchy of automaticity. We observe actions first and words second, and only then do we adjust. And if we are hyper-busy? We may only do the first, which is a major pitfall unless that first is the only reliable piece of data. Take the example of a liar: often, the words and the actions don’t correspond. If you’re too busy to focus on the words but do code the actions as somehow duplicitous, you will end up with a more accurate picture than the more attentive observer. And more broadly, the automaticity of the process is often a good thing. Many times, the person who is acting in an anxious fashion really is anxious and the man who looks reliable to a jury is in fact so. Having these cognitive shortcuts allows us to take information in more quickly and efficiently. It allows us to continue to function instead of getting stuck.

But what about the errors? Here is where a Sherlock Holmes comes in. For him, each of the three parts of person perception has been automatic. He corrects as quickly as he categorizes and characterizes. And that does not have to be a literary ideal. Studies have shown that it is in fact possible to train yourself to correct more naturally, to be less error-prone as a matter of course. It takes practice, vigilance, and a knowledge of the types of errors you yourself tend to make. But all the same, it’s possible. There’s reason to hope.

So the next time you jump to a conclusion about a person, or say with certainty that they are so-and-so or such-and-such, think: why am I saying it? Is there something I’ve forgotten? Something I haven’t applied? Something I may not even have noticed? Then, you may still want to say the exact same thing, but at least you will have stopped for a moment to give it some thought. And maybe, you will have decided to hedge your bets a bit instead of barreling on with full confidence in your own perception.

Photo credit: Holmes and Watson see their clinet for the first time, in “The Adventure of the Norwood Builder.” By Sidney Paget (1860 – 1908) (Strand Magazine) [Public domain], via Wikimedia Commons


Previously in this series:

Don’t Just See, Observe: What Sherlock Holmes Can Teach Us About Mindful Decisions
Lessons from Sherlock Holmes: Paying Attention to What Isn’t There
Lessons from Sherlock Holmes: Cultivate What You Know to Optimize How You Decide
Lessons from Sherlock Holmes: Perspective Is Everything, Details Alone Are Nothing
Lessons from Sherlock Holmes: Don’t Underestimate the Importance of Imagination
Lessons from Sherlock Holmes: Confidence Is good; Overconfidence, Not So Much
Lessons from Sherlock Holmes: The Situation Is in the Mindset of the Observer
Lessons from Sherlock Holmes: The Power of Public Opinion
Lessons from Sherlock Holmes: Don’t Tangle Two Lines of Thought
Lessons from Sherlock Holmes: Breadth of Knowledge Is Essential
Lessons from Sherlock Holmes: Don’t Decide Before You Decide
Lessons from Sherlock Holmes: Trust in The Facts, Not Your Version of Them

Maria Konnikova About the Author: Maria Konnikova is a writer living in New York City. She is the author of the New York Times best-seller MASTERMIND (Viking, 2013) and received her PhD in Psychology from Columbia University. Follow on Twitter @mkonnikova.

The views expressed are those of the author and are not necessarily those of Scientific American.

Comments 3 Comments

Add Comment
  1. 1. gmperkins 9:49 am 10/12/2011

    The final point, to error-check yourself (everyone needs to learn to unleash their inner cynic ;) , is a great one and much needed these days since marketing and politicians really use built in human instincts and implicit mental short cuts to their advantage.

    I recall in a previous article in this series someone saying that this should be taught in school. I definitely agree with that but there seem to be a number of hurdles that need to be overcome first (based on my own attempts at teaching these points to kids, teens and adults). The first is that the age range for when someone is ready appears to be very large. The second is motivation (these kinds of ‘extra steps’ are metally taxing (though it does get easier with practice)). The third problem is when any sort of feelings/emotions are involved because feelings/emotions seem to have a large veto power when attempting to fix a “thought” (even if someone realizes their mistake, they seem to then contain both the mistake thought+emotion and the new conclusion in memory). And there are probably others.

    Like memorization methods (which, sadly, are only taught by teachers who know them, they do not seem to be included in any educational curriculum), it would be great if they were taught at some point (late highschool, even a college course). I think further analysis is needed but a good start would be a class that basically covers Holmes as these articles do.

    Link to this
  2. 2. jgrosay 10:54 am 10/12/2011

    Some say “The face is the mirror of the soul”, and “Beyond 21, a person is responsible of his/her face”.
    How many of the mostly genetically ruled facial features have a connection to traits in how the mind of that person acts ?. The subject can be much more complex than “love at first sight, divorce at first fight”. Salut +

    Link to this
  3. 3. Joseph C Moore, Cpo USN Ret 1:46 pm 10/13/2011

    Thank you, Maria, for your erudition.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article