Illusion of Transparency: Why No One Understands You

In hind­sight bias, peo­ple who know the out­come of a situ­a­tion be­lieve the out­come should have been easy to pre­dict in ad­vance. Know­ing the out­come, we rein­ter­pret the situ­a­tion in light of that out­come. Even when warned, we can’t de-in­ter­pret to em­pathize with some­one who doesn’t know what we know.

Closely re­lated is the illu­sion of trans­parency : We always know what we mean by our words, and so we ex­pect oth­ers to know it too. Read­ing our own writ­ing, the in­tended in­ter­pre­ta­tion falls eas­ily into place, guided by our knowl­edge of what we re­ally meant. It’s hard to em­pathize with some­one who must in­ter­pret blindly, guided only by the words.

June recom­mends a restau­rant to Mark; Mark dines there and dis­cov­ers (a) unim­pres­sive food and mediocre ser­vice or (b) deli­cious food and im­pec­ca­ble ser­vice. Then Mark leaves the fol­low­ing mes­sage on June’s an­swer­ing ma­chine: “June, I just finished din­ner at the restau­rant you recom­mended, and I must say, it was mar­velous, just mar­velous.” Keysar (1994) pre­sented a group of sub­jects with sce­nario (a), and 59% thought that Mark’s mes­sage was sar­cas­tic and that Jane would per­ceive the sar­casm.1 Among other sub­jects, told sce­nario (b), only 3% thought that Jane would per­ceive Mark’s mes­sage as sar­cas­tic. Keysar and Barr (2002) seem to in­di­cate that an ac­tual voice mes­sage was played back to the sub­jects.2 Keysar (1998) showed that if sub­jects were told that the restau­rant was hor­rible but that Mark wanted to con­ceal his re­sponse, they be­lieved June would not per­ceive sar­casm in the (same) mes­sage:3

They were just as likely to pre­dict that she would per­ceive sar­casm when he at­tempted to con­ceal his nega­tive ex­pe­rience as when he had a pos­i­tive ex­pe­rience and was truly sincere. So par­ti­ci­pants took Mark’s com­mu­nica­tive in­ten­tion as trans­par­ent. It was as if they as­sumed that June would per­ceive what­ever in­ten­tion Mark wanted her to per­ceive.4

“The goose hangs high” is an ar­chaic English idiom that has passed out of use in mod­ern lan­guage. Keysar and Bly (1995) told one group of sub­jects that “the goose hangs high” meant that the fu­ture looks good; an­other group of sub­jects learned that “the goose hangs high” meant the fu­ture looks gloomy.5 Sub­jects were then asked which of these two mean­ings an un­in­formed listener would be more likely to at­tribute to the idiom. Each group thought that listen­ers would per­ceive the mean­ing pre­sented as “stan­dard.”6

Keysar and Henly (2002) tested the cal­ibra­tion of speak­ers: Would speak­ers un­der­es­ti­mate, over­es­ti­mate, or cor­rectly es­ti­mate how of­ten listen­ers un­der­stood them?7 Speak­ers were given am­bigu­ous sen­tences (“The man is chas­ing a woman on a bi­cy­cle.”) and dis­am­biguat­ing pic­tures (a man run­ning af­ter a cy­cling woman). Speak­ers were then asked to ut­ter the words in front of ad­dressees, and asked to es­ti­mate how many ad­dressees un­der­stood the in­tended mean­ing. Speak­ers thought that they were un­der­stood in 72% of cases and were ac­tu­ally un­der­stood in 61% of cases. When ad­dressees did not un­der­stand, speak­ers thought they did in 46% of cases; when ad­dressees did un­der­stand, speak­ers thought they did not in only 12% of cases.

Ad­di­tional sub­jects who over­heard the ex­pla­na­tion showed no such bias, ex­pect­ing listen­ers to un­der­stand in only 56% of cases.

As Keysar and Barr note, two days be­fore Ger­many’s at­tack on Poland, Cham­ber­lain sent a let­ter in­tended to make it clear that Bri­tain would fight if any in­va­sion oc­curred. The let­ter, phrased in po­lite diplo­matese, was heard by Hitler as con­cili­a­tory—and the tanks rol­led.

Be not too quick to blame those who mi­s­un­der­stand your perfectly clear sen­tences, spo­ken or writ­ten. Chances are, your words are more am­bigu­ous than you think.

1 Boaz Keysar, “The Illu­sory Trans­parency of In­ten­tion: Lin­guis­tic Per­spec­tive Tak­ing in Text,” Cog­ni­tive Psy­chol­ogy 26 (2 1994): 165–208 .

2 Boaz Keysar and Dale J. Barr, “Self-An­chor­ing in Con­ver­sa­tion: Why Lan­guage Users Do Not Do What They ‘Should,’” in Heuris­tics and Bi­ases: The Psy­chol­ogy of In­tu­itive Judg­ment, ed. Griffin Gilovich and Daniel Kah­ne­man (New York: Cam­bridge Univer­sity Press, 2002), 150–166 .

3 Boaz Keysar, “Lan­guage Users as Prob­lem Solvers: Just What Am­bi­guity Prob­lem Do They Solve?,” in So­cial and Cog­ni­tive Ap­proaches to In­ter­per­sonal Com­mu­ni­ca­tion, ed. Su­san R. Fus­sell and Roger J. Kreuz (Mah­wah, NJ: Lawrence Er­lbaum As­so­ci­ates, 1998), 175–200 .

4 The word­ing here is from Keysar and Barr.

5 Boaz Keysar and Brid­get Bly, “In­tu­itions of the Trans­parency of Idioms: Can One Keep a Se­cret by Spilling the Beans?,” Jour­nal of Me­mory and Lan­guage 34 (1 1995): 89–109 .

6 Other idioms tested in­cluded “come the un­cle over some­one,” “to go by the board,” and “to lay out in laven­der.” Ah, English, such a lovely lan­guage.

7 Boaz Keysar and Anne S. Henly, “Speak­ers’ Over­es­ti­ma­tion of Their Effec­tive­ness,” Psy­cholog­i­cal Science 13 (3 2002): 207–212 .