Want to eradicate racism and sexism? Define them first.

My department chair, Dr. Jeffrey Lieberman, was recently suspended. In a Twitter post, commenting on a photo of fashion model Nyakim Gatwech, who is of South Sudanese descent, he wrote, “Whether a work of art or freak of nature she’s a beautiful sight to behold.”

I wondered, “Why in the hell is he remarking on a photo of a model?!”

I wasn’t alone. Department leaders sent an email to staff and faculty announcing Lieberman’s suspension and “condemn[ed] the racism and sexism reflected” in his tweet.

Lieberman was also ousted from his post as psychiatrist-in-chief at Columbia University Medical Center and resigned as executive director of the New York State Psychiatric Institute. He remains on the faculty at Columbia.

At first blush, university authorities identified a problem and immediately corrected it.

But wait, what was the specific problem? Presumably, it was “racism” and “sexism.” True, his tweet drew on a long history of dehumanizing Black people and objectifying women. Also true, his comments were insulting and hurtful to many, me included.

But university officials didn’t specify the reason for removing Lieberman. Nor did they define the terms racism and sexism. Also lost in the brouhaha is the fact that he has held these leadership roles for 17 years. If racism and sexism are, in fact, the reasons for his dismissal, then administrators should conduct an investigation to determine what impact, if any, his leadership has had on subordinates, hiring and promotional practices, among others.

I am concerned that these glaring omissions will facilitate the administration’s quick “resolution” of this incident without actually understanding the scope of the potential problem and how to correct it.

Meanwhile, not only are we left to speculate about why Lieberman was dismissed but also what constitutes racism and sexism. Is it the mere use of language that dehumanizes a person, regardless of intent? If so, then people who describe talented Black athletes, for example, as “beasts” — a not uncommon term on sports radio — are racists, right? Or is it the use of objectifying language to describe persons from genders other than one’s own?

This trivializes racism and invites disingenuous criticism of “political correctness” and “wokeism” among intellectual elites in America.

“Implicit bias” is often used euphemistically to avoid accountability for racism. This is wrong for many reasons, not least because unconscious attitudes may or may not play a role in behavioral acts of racism. Simply having an implicit bias does not mean a person will inevitably act on this bias. Nor does it mean a specific act of racial discrimination was due to such bias. Focusing on implicit bias — on a person’s thoughts, rather than on that person’s harmful acts — tends to obfuscate the issue. That’s why authorities are obliged to delineate the actual, observable behaviors that meet their criteria for racism and sexism. The key is to stay focused on people’s behaviors, rather than to speculate about their motives. It’s impossible to know, for certain, what’s in a person’s head or heart.

In my research, I define racism as an action that results in disproportionately unjust or unfair treatment of persons from a specific racial group. For instance, Black people are far more likely than their white counterparts to be arrested for drugs, even though both groups use and sell drugs at similar rates.

Does that mean police who enforce drug laws — or legislators who pass them — are racist? Not necessarily. First, we would need to assess the person’s (or institution’s) response to reasonable evidence that their actions contribute to racism. If they fail to take corrective steps after being presented evidence indicating their participation in racism, then the label “racist” is appropriate. On the other hand, if one unwittingly participates in racial discrimination but corrects the offending behavior when the discrimination is brought to their attention, it would be inappropriate to label the person a racist. We all make mistakes.

For example, 16 of 20 Congressional Black Caucus members supported the Anti-Drug Abuse Act of 1986 (a.k.a. crack-powder law), setting penalties that were 100 times harsher for crack than for powder-cocaine convictions. It was discovered that more than 90% of those sentenced under this law were Black. In 1988, only eight of 22 CBC members supported extending the law. Members who changed their vote wouldn’t qualify as racist.

Lieberman realized his tweet was inappropriate hours after posting it — but before his suspension — and initiated damage control. In an email sent to colleagues, he apologized for posting “a message that was racist and sexist” and promised to make “needed personal changes.”

I don’t know if Lieberman is a racist; I would need more information. I do know that in response to a New York Times review of my book, “Drug Use for Grown-ups,” Lieberman sent an ominous email to more than a dozen of my non-Black colleagues. He appeared to be fishing for some sort of negative consensus regarding my book’s contents. It was disappointing but not surprising that he didn’t contact me directly nor include me on his email. When I contacted him, he ultimately apologized.

I hope the university takes full advantage of this teachable moment. Addressing questions about whether Lieberman has or continues to engage in behaviors unevenly harmful to Black people or women — while clearly defining what constitutes racism and sexism — would be a good start.

Hart is Ziff Professor at Columbia and author of “Drug Use for Grown-ups: Chasing Liberty in the Land of Fear.”