I just stared at a student, stunned when I heard him say none of my students are bright. All I could mutter was, “I teach plenty of bright kids. All of my kids are bright. They’re brilliant.”

He dropped his pile of textbooks on my desk and took my load from me. “It’s just… you know. They’re not as smart as we are.” He shrugged, turned, and waved goodbye to me, heading off down the hall to his next class. I stared at my empty doorway, my mind spluttering out failed retorts and constructing a more successful “teaching moment” for the next time the subject came up.

And it did. Since that one interaction with a student (who I happen not to teach), I’ve become increasingly aware of the way in which we subliminally message our perceptions of, and our confidence in, our students.

I work at a school that divides its students up into different academies: ASE, the Academy of Science and Engineering, and NSST, the Academy of National Safety and Security Technologies. One has designed its curriculum around inquiry, partnering with Vanderbilt scientists to expose its students to rigorous content and engaging experiments. The other, the newer of the two and by far the largest, is becoming more and more successful at weaving in themes of criminal justice and technology into all its courses, teaching an increasingly relevant perspective of the world to a new generation of Americans.

Both academies offer great opportunities to students; both have the potential to teach and inspire.

But I’ve noticed a trend. Perhaps because NSST is the larger of the two, perhaps because ASE students are privy to certain exceptional opportunities, my students perceive one academy as “smart” and the other as the one full of the “dumb kids.”

I’ve also noticed this trend when listening to fellow teachers and administrators—even when hearing the words that come out of my own mouth. When talking about NSST, we talk about behavior. (“You’ve all done a great job with your discipline today, students. We are very proud of you!”) When we talk about ASE, we talk about ACT SCOREs, science projects, and field trips to colleges.

How must this affect my students, the NSST kids? What damage are we doing to their self-images, their self-confidence, and, subsequently, their academic success? According to research as far back as the 1960s, potentially quite a bit. In 1964, Harvard professor Robert Rosenthal conducted an experiment on elementary students in which all kids were given a “special” IQ test. Afterwards, the researchers identified a randomly selected group of those kids as the “special” ones—the ones that were about to show significant and extraordinary intellectual growth. They shared the names of this group with the students’ teachers.

Over the next two years, Rosenthal found that this randomly selected group of students generally performed better than their peers. “If teachers had been led to expect great gains in IQ,” he explains, “then increasingly, those kids gained more IQ.”

Today, while researchers work to better understand this effect and strategies to keep teachers from falling into the trap of low expectations for certain students, we need to start by recognizing the risk ourselves—getting comfortable with the knowledge that our carefully planned lessons and curricula are only one small piece of the message we send to our students. And with so much recent (and worthwhile) focus on teaching character traits through deliberate messaging (I am thinking, particularly, of the kind of planning and messaging detailed here), we have to realize that our casual comments affect our kids just as much, if not more, than our buzzwords and mottos.

If we truly have a vision for our kids to be a certain way (patriotic, curious, confident, intelligent, etc.), we have to live and speak as if we believe it with every word. Everything we do or say, whether planned or unplanned, has to drive them towards that goal. Otherwise, how many more students will I have to try to convince that I, too, teach “bright” kids?