A hand draws a heart on an apology card using a pink crayon.

Responsibility & Four Levels of Apology

Apologies can be crucial for forgiveness and reducing anger (McCullough et al 2014). It took my partner and I a long time to realize that we did not understand ‘apology’ in the same way. Eventually, we agreed on some distinctions. And they’ve been pretty useful. In this post, I’ll summarize how we distinguish degrees of responsibility and severity.

1. Four Levels of Apology

My partner and I like speaking precisely. “Precision of language!” we tease one another (Lowry 2002). To make our apologies more precise about responsibility, we employ some distinctions.† Depending on how responsible we feel, we will offer one of these four levels of apology:

  1. Compassionate: “I’m sorry about X even though X is not, in any relevant way, my fault.” For instance, I might say “I am sorry to hear about your grandmother’s passing” even though I played no part in your grandmother’s death (Lazare 2005, Chapter 2).
  2. Agnostic: “I’m sorry about X, and I think it’s possible that I am responsible for X, but I honestly don’t know whether or not I’m (even partly) responsible.” This is similar to a conditional apology: “If I am responsible for X, then I am sorry.”
  3. Partial: “I’m sorry about X and I take myself to be at least partially responsible for X.” In this case, I was only part of the event in question. Other factors also contributed to the event.
  4. Full: “I’m sorry about X and I’m (for our intents and purposes) fully responsible for X.”

2. Four Levels of Badness

In the heat of an argument, we can easily lose perspective. So sometimes, after I have apologized and reconciled with my partner, we find ourselves asking, “How big a deal was that, really?”

There could be a couple things at play in this question. First, this question might be about whether some action was really as bad as it seemed initially. Second, the question might be about whether one of us really made some sort of error. With these two variables in mind, an answer to our question begins to emerge.

Bad? Error?  Big deal?
✔︎ Negligible Probably not
✔︎ Unfortunate Potentially
✔︎ ✔︎ Human
VERY! YES! Inhuman Probably

The Four Levels

Perhaps you want to know more about what I mean by ‘negligible’, ‘unfortunate’, ‘human’, and/or ‘inhuman’?” in the table above. That’s fair. Here’s the rough draft of what I have in mind.

  • Negligible. I made an error, but it’s not a big deal. It doesn’t have substantially undesirable consequences for anyone. In fact, making a fuss about the error is probably a more serious error than the error itself.
  • Unfortunate. I caused something non-negligibly bad, but what I did wasn’t really an error. Imagine, for instance, that you do something which contributes to a very undesirable outcome. However, you were acting under conditions of uncertainty and what you did was, in fact, optimal. In this case, you have not made an error. So, the bad outcome is merely unfortunate.
  • Human. I made an error and it’s non-negligibly bad. However, the error is entirely understandable (even predictable) in terms of what science tells us about humans (e.g., we’re biased, we are prone to certain mistakes, we get hangry, etc.). (NB: since it’s predictable, there might be ways to prevent or inhibit it.††)
  • Inhuman. We simply cannot make sense of my error; it’s that erroneous. And what I did is so bad that …well, it’s pure, undiluted evil.†††

3. An Example

These days, my partner and I use the first set of distinctions pretty much every time we apologize.

Nick: I’m sorry honey.

Hannah: What level?

Nick: Level 3.

And we can add further precision to the conversation with the latter distinctions.

Hannah: Thanks honey. I guess it’s not a big deal. It’s just unfortunate.

According to the distinctions, I’ve apologized and admitted that I think I’m at least partially responsible. And Hannah humbly admits that I haven’t, strictly speaking, made an error. My apology assures Hannah that I care. Hannah’s response reminds me that I do not need to get defensive.

4. Conclusion

Those are the distinctions. We distinguish levels of responsibility and the likelihood that something is a big deal. So far these have been helpful.



† And, no: I don’t think free will is necessary for responsibility. (And, if I am honest, I am agnostic or skeptical about free will, depending on the day.)

†† I am inspired here by the forward-looking approach to action and responsibility from Gregg Carusso (2013). The book is below, but here are some more digestible forms of the view: TEDx talk, interview on Philosophy Bites podcast, interview on Informal Hour podcast.

††† To be honest, I’m not convinced that the ‘inhuman’ category is real. I get the sense that all human error — no matter how severe — will make sense in some (albeit twisted) way. And although I can imagine something that is extremely undesirable, I just can’t imagine something that is pure evil. So even if someone’s behavior is very erroneous and very bad, it would still wouldn’t be inhuman is the sense I have in mind. But suppose that I’m wrong about this and the aforementioned notion of ‘inhuman’ is totally realistic. Wouldn’t the difference between ‘human’ and ‘inhuman’ be a difference in degree, not kind? Also, you might be wondering why I included ‘inhuman’ if I have so many reservations about it. I included it to account for some peoples’ intuition that some actions are “pure evil” — whatever that means — or their intuition that certain behavior merits death, torture, revenge, etc.




Published by

Nick Byrd

Nick is a cognitive scientist at Florida State University studying reasoning, wellbeing, and willpower. Check out his blog at byrdnick.com/blog