That’s not what I meant!

Ambiguity pervades language. This ambiguity can be used strategically by speakers, but it’s also what makes language so challenging for machines to understand – and in some cases, it even leads to miscommunications between people, particularly over written communication.

During in-person interactions, ambiguity is more easily avoided. If a speaker says of a recently released film, “Truly an Oscar-worthy movie”, you can infer whether they’re being sarcastic or not by the tone of their voice, or perhaps whether they accompany the statement with an eye-roll. (You might also simply know the speaker – and the film – well enough to infer whether it’s the kind of film they’d actually love, which is also a relevant piece of information.)

So what happens in a digital age, when so much of our communication is reduced to the written word (email, text, etc.)? Anecdotally, it certainly seems that text messages are very susceptible to misinterpretation (see this Key and Peele video for an entertaining illustration of this).

But if speakers (or writers, technically) are aware of this susceptibility, shouldn’t they take pains to avoid misinterpretation?

The Study

A 2005 study (Kruger et al, 2005) asked this very question. Specifically: why do people write ambiguous emails?

The authors asked participants to write serious and sarcastic statements about different topics (i.e. movies that had come out recently). These participants were then asked to predict how successfully someone else would interpret which statements were sarcastic, and which were not. Later, another group of participants read these statements, and judged them as either sarcastic or not.

As one might expect, readers were not 100% successful in accurately deciphering statements – on average, 84% of the original statements were correctly judged as either sarcastic or not. Perhaps more surprisingly, the first group of participants seemed to be overconfident in the interpretability of their emails, predicting a mean of 97% accuracy.

A second study then contrasted predicted vs. actual accuracy between written and spoken communication. This time, there was an even larger gap in the email condition between expected and actual results (see the figure below); notably, this gap was considerably smaller in the spoken condition. That is, people are in general overly confident that others will be able to interpret what they mean, this tendency is exacerbated – or, rather, inadequately corrected for – when people communicate over email.

Screen Shot 2018-02-05 at 4.30.28 PM.png

Figure 1:Taken from Kruger et al (2005). Participants were less successful in anticipating the accuracy of recipients’ interpretations when the message was conveyed over email vs. voice.

The authors then report the results of several more studies, and conclude that this gap is the result of egocentrism on the part of the email-writer. When participants were forced to confront the ambiguity of their statements (i.e. by being asked them to read them aloud with an alternative interpretation), their predictions of a recipient’s accuracy in judging their intent significantly decreased. So while people are capable of recognizing that they use ambiguous language, but don’t necessarily do so by default.


The Takeaway

Miscommunication over email occurs because of at least two intersecting factors:

  • Reduced information that would otherwise indicate the email-writer’s intentions (tone of voice, gesture, etc.);
  • Inadequate perspective-taking on the part of the email-writer, who fails to account for this reduction in information.

The first obvious question these conclusions raise is why email-writers consistently fail to account for how ambiguous their emails are. Other work (Epley et al, 2004) has suggested that people “anchor” their prediction of how someone else will interpret an utterance to their own interpretation; in other words, the email is already clear to the email-writer, so they don’t stop to think about how it could be misinterpreted. Here, it seems like asking questions about individual differences could better address the underlying mechanisms – do different people vary in how much they adopt the perspective of a potential email recipient? If so, what other dimensions correlate with this variability?

Second, and perhaps more relevantly to our lives today, the emails in this study only involved text. In fact, participants were explicitly instructed to refrain from adding emoticons like “;)”. But now, in a world full of emoticons, abbreviations, and even talking Emojis, email-writers (and texters, etc.) have access to many resources besides the written word. How do people use these resources to reduce the ambiguity of their messages? And are there ever situations in which these other resources actually lead to more misinterpretation (such as receiving an ambiguously-intended “;)” from a colleague)? The world of digital communication offers exciting new innovations in communicative styles, ripe for scientific investigation – but perhaps people will always be prone to miscommunication.


Kruger, J., Epley, N., Parker, J., & Ng, Z. W. (2005). Egocentrism over e-mail: Can we communicate as well as we think?. Journal of personality and social psychology, 89(6), 925.

Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective taking as egocentric anchoring and adjustment. Journal of personality and social psychology, 87(3), 327.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s