Last Thursday, for Cinco de Mayo, Republican presidential candidate Donald Trump tweeted an image of himself eating a taco bowl with the words: "Happy #CincoDeMayo! The best taco bowls are made in Trump Tower Grill. I love Hispanics!"
People were quick to point out some ways in which the tweet was not only inaccurate (taco bowls are available at the Trump Cafe, not the Trump Tower Grill), but also offensive.
For one thing, taco bowls aren't Mexican but rather an American invention (as is, to a large extent, the celebration of Cinco de Mayo itself). For another, Mexican-Americans are just one subset of "Hispanics," and not all Hispanic cultures identify with tacos or Cinco de Mayo.
But there's a deeper reason the tweet might tweak a nerve, and it has to do with the subtle inferences we draw from people's use of language — in this case, the statement: "I love Hispanics!" It's positive on the surface — but the implications aren't all rosy, especially given the context of Trump's positions regarding Hispanic immigrants. In particular, the claim presupposes that it's meaningful and informative to classify people as "Hispanics" when making generalizations about whether they're lovable. And that's a potentially strange (and offensive) view to hold — is there really something that Hispanics all have in common by virtue of which they are (or aren't) lovable?
If you think this analysis credits the critics with too much sophistication in their response to a casual tweet, think again. Research in psychology confirms that people are sensitive to subtle assumptions embedded in the way we talk about social groups, with negative implications sometimes lurking behind superficially positive claims.
As a first example, consider positive stereotypes — generalizations about social groups that attribute a positive attribute to the group. An example could be "Asians are good at math," or "Mexicans make good food." Research suggests that these positive claims can do more harm than good: They lead people to expect that the person who endorsed the positive stereotype also holds negative stereotypes, and they lead people who belong to the stereotyped group (but not necessarily others) to judge the person less likable and more prejudiced. What this suggests is that non-Hispanic Trump supporters might interpret Trump's tweet in a positive light, but it could lead people to believe he's more likely to endorse negative generalizations about Hispanics as well — and it might lead people who identify as Hispanic to judge Trump even less likable than they already tend to find him.
As a second example, consider comparative claims about social groups, such as "girls are as smart as boys." While the content may seem positive on the surface — explicitly claiming equivalence — the linguistic construction implies an asymmetry, with the subject (girls) defined with respect to a reference point (boys), implying that the reference point is more representative of the attribute in question (intelligence). In fact, research suggests that adults respond to such claims of equivalence by assuming that members of the reference category (boys) are naturally disposed toward the attribute, while members of the subject category (girls) have to work harder. As with positive stereotypes, a superficially positive claim carries subtle — but powerful — negative implications.
So people are probably right to question Trump's taco tweet — not only because they doubt his sincerity, but also because even a superficially positive claim can carry negative implications. That said, I mean no offense to Trump. I love billionaires!
Tania Lombrozo is a psychology professor at the University of California, Berkeley. She writes about psychology, cognitive science and philosophy, with occasional forays into parenting and veganism. You can keep up with more of what she is thinking on Twitter: @TaniaLombrozo