Bryan Caplan  

The Diction of Social Desirability Bias

Keep it simple, and pay attent... The Case Against a Basic Incom...
"Sorry, I can't come to your party."  This common excuse is almost always literally false.  You're working?  Unless your boss chains you to your desk, you can come to the party.  You're in Paris, and the party's in DC tomorrow?  If you can beg, borrow, or steal airfare, you can come to the party.  The same goes for most social uses of the word "can't" - everything from "We can't be together" to "I can't help myself."

Why say, "I can't" when the truth is "It's too costly for me" or "I don't feel like it"?  Because "I can't" sounds better.  It insinuates, "The only reason I'm not doing X is because I lack the ability to do X.  Otherwise I would totally do it."  "It's too costly for me" and "I don't feel like it" are insulting by comparison.  Both blurt, "X simply isn't my top priority.  Get used to it."  In short, the way we use the word "can't" is a clear-cut case of Social Desirability Bias: our all-too-human propensity to lie when the truth sounds bad.

The literally-false "can't" is hardly alone.  Social Desirability Bias permeates our diction - i.e., the specific words we choose to use.  Consider the following expressions:

1. "I'll do my best."  Unless you devote 100% of your resources to success, you haven't really done your best, have you?  But it sure sounds nice.  The same goes for "We're doing everything in our power," "I'll stop at nothing," and the like.

2. "We have no choice."  Unless there is literally only one thing you are capable of doing, you have a choice.  So why claim otherwise?  Probably because you're doing something that seems wrong, and you don't feel like justifying your action as the lesser evil.  The same goes for "We were forced to do it" and "I simply have to do this."

3. "Nothing is more important to me, but..."  If nothing is more important, you will sacrifice everything else you have to get and keep X.  So unless you've given your all for X, you're overstating it's importance.  As Social Desirability Bias predicts, X is generally something high-minded: God, country, and family top the list.  "I'll pay any price for X" and "You can't put a price on X" work the same way.

4. "X is unacceptable."  People are capable of accepting almost anything; where there's life, there's hope.  But they'd like much much more.  People who put the "unacceptable" label on an unfavorable deal are covertly bargaining.  Why?  Because it's less confrontational to plead, "I'd take your offer if I could.  But my hands are tied" than "Show me the money!"

The diction of Social Desirability Bias fits neatly with my skeptical take on addiction - and mental illness generally.  To my ears, "I can't stop drinking" directly parallels "I can't come to your party."  Of course you can refrain from drinking alcoholic beverages.  And of course you can attend my party.  But if you declare, "I'd rather stay home and drink alone," you sound bad and listeners get mad.

In Moliere's The Misanthrope, Philinte insists,
In certain cases it would be uncouth,
And most absurd to speak the naked truth;
With all respect for your exalted notions,
It's often best to veil one's true emotions.
Wouldn't the social fabric come undone
If we were wholly frank with everyone?
Philinte's right as far as he goes, but he misses a deeper issue.  When we let Social Desirability Bias rule our diction, there's a grave danger our literally false words will corrupt our thinking as well.  Indeed, such corruption is all around us.

COMMENTS (25 to date)
Robinson writes:

Have you read Scott Alexander's What Universal Human Experiences Are You Missing Without Realizing It? I think you may be missing out on a human experience that affects your perspective.

Suppose you were talking to someone with congenital insensitivity to pain. They said, "I'm skeptical about pain. I mean, I understand why you say 'Ouch!' when you're cut: you worry about blood loss and infection, just like I do. But you want everyone to know that you are wounded, to gain sympathy. 'Pain' is your word for this social phenomenon." Surely you'd agree they didn't understand pain.

I have a suspicion you don't experience akrasia or executive dysfunction (and I'm pretty confident you haven't been addicted to anything). There's a fairly common human experience of "I know I shouldn't be doing this, but I'm going to" (whether it's a habit as destructive as heroin or as mundane as reading blogs at work) characterized by immediate regret and self-loathing. Modeling it as "you decided you preferred to take Action A than Action B" is like modeling the experience of pain as "You preferred that your skin had not been cut, because it could lead to infection."

To be clear, your lack of akrasia is a virtuous personal trait on your part! I'm guessing it plays a role in your high productivity and discipline. It's even understandable that you might find people on the other side of the spectrum unpleasant and keep them out of your bubble. But I really do think when you imagine their minds as working like yours, you're missing an important part of the model.

Cliff writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Ally writes:

"I can't" is often useful shorthand. I think most people recognise that it usually isn't literally true. Except where something is physically impossible, or we lack the means to achieve it.

Similar observations have been made regarding use of the phrase "I haven't got the time..."

Everyone has the same 24 hours to use each day.

When people say "I haven't got the time to do X." what they usually mean is "X is not a priority for me right now.", but this can often sound rude as it makes the person it is directed toward feel unimportant.

Matt Skene writes:

Consider the following argument:

When people say they are too sick to work, this is almost never literally the case. Even if you have a very high fever and are constantly needing to throw up, you still could make it into work and put in some productive efforts toward your job. What people really mean is that their physical condition makes it more costly to engage in certain activities, and that due to the increased cost, they are unwilling to do them. However, since they still could do them, when someone says "I'm too sick to work" all they really mean is "I'd rather stay home than put in the effort to be productive today." Since staying home is merely an expression of their preferences on that day, we shouldn't really think they were sick in the first place, and the whole idea of health is therefore suspect.

This appears to be your argument against mental illness. It's true that a mental illness consists in is a condition that shifts the cost structure of activities for individuals in such a way that they consistently are unwilling to pay the cost of doing activities that are usually expected of people. But this is true of all illnesses. Given this, I don't see how the fact that you literally could still do those things can be evidence for the non-existence of mental illnesses unless it's evidence for the non-existence of illness in general.

Greg G writes:

I always find it amusing and ironic when libertarians complain about some of the most common and popular language conventions. Nothing could be more libertarian than the way language conventions develop. There is no central authority on language. Dictionaries change annually based on the voluntary language choices made by millions of individuals.

Most people understand perfectly well that there are many language conventions and figures of speech that almost no one intends 100% literally. For them, there is no " grave danger our literally false words will corrupt our thinking as well." Most people understand that the literal meaning is often different from the actual meaning and the difference can usually be chalked up to courtesy.

It is a good thing that most people speak in a more diplomatic and less confrontational way than a strictly literal word choice would produce. These cases aren't examples of "lies" or "corruption." They are cases where people use common sense and common conventions to reduce friction and hostility in human interactions.

You might think it would be a good thing if people were more literal and "insulting" in their everyday speech. It wouldn't be.

I often find myself arguing on these forums that some pie-in-the-sky libertarian idea wouldn't turn out as well as libertarians think in some counterfactual. Well, language is the current, actual human practice most determined by libertarian principles....and no one seems more unhappy with the result than libertarians who are constantly telling us that conventional language uses are wrong.

Daniel writes:

You assume that the listener is somehow being fooled by literal language here. They are not, if they have experience with basic social norms. Everyone knows that when someone says "I can't, because...", they mean "I can't reasonably make it without sacrificing very important things, such as..." The same goes for all of your other examples. "I'll do my best" usually means "I'll do an optimal job given the constraints of ongoing, responsible human life." Using ordinary language that it isn't meant to be understood *literally* isn't some great character flaw.

And for #3, you've got it wrong: "If nothing is more important, you will sacrifice everything else you have to get and keep X." No, that's only true if *everything combined* is less important than X.

Curt Doolittle writes:


You're confusing logical(independent of cost) statements, with rational(considering cost) statements.

Humans evolved to consider costs, whether, detecting the imposition of costs (morality) or the return on cooperation, in all statements. Humans speak RATIONALLY not LOGICALLY.

Which, for an economist, is a very .... well, strange if not unbelievable error.


Curt Doolittle
The Propertarian Institute
Kiev, Ukraine

Ben H. writes:

Robinson, bingo. What I find most problematic about Bryan's position on mental illness is that he jumps from "this is how I personally experience reality" to "this is how I believe everyone experiences reality" without confronting the obvious problems in making that leap. Maybe, just maybe, one ought to trust the self-reported testimony of the legions of the mentally ill who say that their experience of reality is different from what Szasz claimed. (Not to mention the considerable evidence from science that Szasz was wrong – I posted a link on Bryan's FB wall just yesterday with one such study – but that is another issue, for another day.)

Tiago writes:

Robinson and Ben H., agreed.

For such a prolific thinker, this seems to be the one important blind spot I see in Bryan.

Thomas Sewell writes:

I'm sorry, I'll do my best, because nothing is more important to me, but having no choice in the matter, I can't tell you how unacceptable this comment is.

How's that?

Jameson writes:

Greg G's comment is great. It's sad, too, because classic thinkers like Milton Friedman and Friedrich Hayek specifically praised the development of language as an example of the kind of unplanned order they were talking about. Personally, I find today's libertarian bloggers far too rationalistic, in the sense that Hayek would have sharply criticized.

To be fair, it's true that language can become poisoned (as Hayek so forcefully argued in The Fatal Conceit). But I don't think this is one of those cases.

When I tell my wife she is the most beautiful woman in the world, somewhere in my distant subconscious I can hear Professor Caplan saying, "But that's just not literally true!" And in my distant subconscious I softly but firmly respond, "Shut up, Bryan."

Philo writes:

"X is unacceptable" is probably short for, "Under the circumstances, X is unacceptable": in other words, "As things stand there is a pretty good chance that, merely by demanding it, I (we) would get something better than X." Of course, this is often said when it is not true, and not even believed to be true; so there may still be Social Desirability Bias, or maybe just Wishful Thinking, in play.

Philo writes:

@ Greg G, on libertarianism and language

Bryan isn't trying to exercise centralized authority over language; he's not even arguing against the white-lying, spinning, soft-pedaling, etc., that we all routinely do. He is proposing that we not let these routine literally false utterances affect our own *thinking*. "Clear your *mind* of cant" (Dr. Johnson).

Jeff writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Jesse C writes:

I think some others are reading too much into this post. I think the last paragraph makes a good point, depending on the degree implied by "grave."

One example of language where the literal meaning definitely does corrupt our thinking is the overuse of "priority" when describing relative tradeoffs. E.g. everyone agrees that human life is more valuable than a stick of gum but nobody suggests we outlaw all gum (and ALL similarly frivolous things) to spend more on ambulatory services, etc, until we've exhausted our efforts at protecting life. THAT is a trade-off we all accept but no politician can describe it as such.

(I believe the example comes from Thomas Sowell, albeit butchered)

Segeant Popwell writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Greg G writes:


Of course you are right that Bryan is not trying to exercise centralized authority over language. I really don't think I suggested he was.

I was challenging his claim that our thinking really is corrupted by common phrases that everyone understands aren't literally true.

And I was challenging the idea that the world would be a better place if everyone spoke in the more confrontational and less diplomatic way that would result from strictly literal "uncorrupted" word choices.

Henri Hein writes:

Whenever I say "I can't make it," I always feel bad about lying to my friends. Of course I could make it. I just prioritized something else higher.

Bryan's point is emphasized by the fact that whether cancellation is followed with an explanation depends on what that explanation is. If you would rather just stay home and watch Netflix, then you will just say "sorry, can't make it." If your wife is in the hospital having a baby, then that is a perfectly acceptable explanation for not joining your co-workers at happy hour.

Greg: The idea is not to change people's behavior. It is to warn all of us about the danger of diction confusing our thinking. It is a point well taken.

Cliff writes:

Yes, obviously common shorthand that is understood by everyone lends strong support for your position that severe mental illness does not exist.

All depressed people just prefer to lay in bed all day and then kill themselves? It's just a coincidence that the drugs improve their condition substantially?

It's impossible that ADHD people have more trouble concentrating because of something other than having a strong preference for flitting from one activity to another?

Anti-psychotics operate by surgically removing your preference for hallucinations?

Apparently moderators will not allow Thomas Szasz to be criticized (who Caplan links to approvingly), so I won't.

[A critique of someone's ideas is fine on EconLog. A critique of someone's background--an ad hominem--is not allowed on EconLog. See our longstanding, clearly stated comment policies. --Econlib Ed.]

Jameson writes:

All of the commenters trying to defend this post are reducing its meaning down to nothing. If those of us who are critical are "reading too much into it," then would somebody just tell me what Bryan is literally asking us to do? Come on, just give it to me straight. No social desirability bias now...

Because it really sounds like he wants us to stop saying things like, "Sorry, can't make it." I am not going become a jerk in my ordinary life just so I can think more objectively about political and economic issues. In fact, I'm pretty sure it wouldn't help my political philosophy in any way.

MikeM writes:

Another one is "this is ridiculous". The social effect is the opposite though, you're trying to overstate your case. "This return policy is ridiculous!" is a common refrain from upset shoppers but a moment's reflection about profit maximization will make one realize that the store policy is neither illogical or ludicrous.

Greg G writes:

>---"Sorry, I can't come to your party." This common excuse is almost always literally false..."

A much more honest answer might be: "Your party will probably be a terrible drag. There are dozens of things I'd rather do. I never really considered going."

One of these possible responses does indeed reflect gravely muddled thinking about the best way to respond. Not everyone agrees on which one that is apparently.

AnonymousUser writes:

I guess I had just the preference to worry about various things and my self-worth when I'm trying to sleep and I was just trying to fool someone because you suppose to sleep during night. I'm still not quite sure who that might be as I'm single.

I'm now convinced that I was forced by society to seek help and by Big Pharma to start taking medicine. Then I found the strength to start being really convincing that I'm not worried over everything. I am not sure how society was forcing me as I was trying to hide it but now I'm sure it was trying to force me somehow.

Sorry - the case you are trying to make is very remote from vast majority of those 20% and based on simple stereotypes that it's closer to insensitive then insightful. If the people just use mental illness as an excuse then it would make no sense to try to self-medicate or otherwise seek help in secret - which is what people do given the fear of stigma.

Of course we can model it as preference to change preference and economics will be still fine with it but I don't think leap to "medical illness is fake" is warranted - it seems like misapplication of economics.

Anon writes:

What do you know about alcoholism? Have you done ever anything that requires such emotional shift?

I remember reading an interesting article about addiction and Native Americans. Apparently NA's used alcohol fine for generations but when they were ripped off from their societies and feeling of community, the alcoholism started. Same thing for war vets and PTSD.

When you are looking at an addiction you are just looking at preferences and this is exactly why nerds suck at social science. I know because I used to think exactly the way Bryan did.

Have you ever been in a support group?

For example dropping weight isn't just about calories. It's about identities, its about personal stories. Sure its a preference to slacking and overeating over exercises and diet but this does not help the emotional wall such people face.

But yes I suggest everyone to do honest introspection and take responsibility.

J. Goard writes:

I'm both a (sort of moderate) libertarian and a linguist, and I'm with Greg G. On the whole, Bryan's examples aren't social lying, but rather run-of-the mill pragmatics. As malevolent genie stories teach us, it's a basic property of the form-meaning mappings in language that they require a contextualizing mechanism to be interpreted at all. Even the absurd interpretations being described as "literal" are not free of a pragmatics, they just have a different one.

Comments for this entry have been closed
Return to top