Sorry

June 15, 2015

In April 2014, my partner at the time, Ad, mentioned that he thought I apologized too often and too unnecessarily. While neither of us thought this was necessarily problematic (I'd rather apologize too much than not enough), I disagreed with his assertion. Data is usually the most reliable way to settle disagreements, so we decided to track apologies to gain a better understanding of this aspect of our behavior.

Since most of our communication was documented in chat logs (long-distance relationship), we were able to identify most of our apologies by querying our conversations from August 2013 (when we started talking as friends) to April 2014 for the words "sorry" and "sry." This tactic excluded other forms of apologies, like "my bad," but we figured it would capture the majority of our apologetic sentiments. I took the output of the query, analyzed it, and stuck it in a shared Google spreadsheet:


We also did live tracking of apologies from April to June. Our goal was to see if we could have a "no sorry" week, so we were actively trying to mitigate our apologies. We continued to log only IMs and text, excluding apologies that we made vocally.

When everything was said and done, there were 455 total apologies. Ad apologized 266 times (58%), and I apologized 189 times (42%). Here's how it broke down over time:

Conversation and Apologies throughout Time

Ad and I met in person and discovered our chemistry in December, which is why you see the spike in messages. We were extra polite during the first few months that we were getting to know each other, and then the number of apologies moves in reasonable proportion to the amount of conversation we were having, up until April/May when we decided to try to apologize less. (The dip in conversation in March was caused by more time than usual spent in person.)

That's all peachy, but what sorts of things were we apologizing for?

Apologies by Category

Availability apologies were related to our ability to be present (both mentally and schedule-wise) and give each other attention.

• A: "sorry i was gone so long"
• R: "sorry/ phone died/ at mom's now"

Behavior apologies were related to the way our actions impacted each other.

• A: "sorry if i'm putting you on the spot"
• R: "sorry I am not being very helpful"

Communication apologies were related to how effectively we conveyed information.

• A: "sorry about the misunderstanding"
• R: "got confused about timezones, sorry"

Sympathy apologies were related to times we expressed remorse for unfortunate things happening in the other person's life.

• A: "sorry to hear about your grandmother"
• R: "I'm sorry you've been feeling sicky lately"

Joke apologies included the times that we said "sorry" or "sry" in a sarcastic or insincere way.

• A: "SORRY TO TEAR DOWN YOUR SPECIALNESS HYPOTHESIS"
• R: "SORRY EVERYONE ELSE [BUT ME] IS LAME"

Personality apologies were related to the ways that we felt our preferences or feelings were in conflict with the other's preferences or feelings.

• A: "ack, sorry about that. I'm SO awkward on the phone"
• R: "sorry for bein weird!"

???? apologies comprised of apologies whose purpose I could not discern with the context I was given, and were mostly excluded from analysis.

Now that we know what we were apologizing for, we can see who apologized more for what.

Apologies per Person per Category

Remember how Ad apologized a whole lot more than I did in the time we were tracking? Rather than apologizing 40% more than me consistently across all categories, the difference is almost entirely caused by his apologies for availability and behavior. We were pretty evenly matched on everything else, except I apparently think apologies are funnier than he does.

What sorts of factors could cause the huge difference in, say, apologies related to behavior?

1. Maybe Ad is particularly sensitive to the way his behavior affects other people
2. Maybe I am particularly insensitive to the way my behavior affects other people
3. Maybe Ad is an asshole and has a lot to apologize for
4. Maybe I'm an angel and never do anything that requires apology
5. Any combination of the above (and lots of other possible reasons)

Obviously we can't claim causality for anything in this analysis, but it's still interesting to think about.

If we look at types of apologies over time, we can see some revealing things:

Ad's Apologies over Time

Ad was mostly equally apologetic about everything until we met in person in December. After that, he apologized more for the way he was treating me and the attention he was able to give me, even though I don't believe I ever asked him to be more available. (It's interesting how expectations for how we should treat each other change!)

My time-based apology trends are a little more convoluted:

Robin's Apologies over Time

I was pretty unapologetic during the months we were explicitly friends, but I became sensitive about my availability earlier than Ad became sensitive about his. There was a big spike in my apologies for communication in December and January, which is when we were trying to establish some relationship expectations and figure out what we could be for each other - the amount of effort I was putting into communicating clearly shows. There's also a big leap in availability apologies in February, but it was kind of an anomaly for no good reason. Finally, there's a noticeable jump in sympathy apologies in January. It was a bad month for Ad - one of his friends passed away, there was stuff going on with his family, and he was facing some difficult career decisions. My sympathy was palpable (or rather, his misfortune was).

Just in case you were wondering if our message-based apologizes were any different on and off the road...:

IM vs Text Apologies

As a whole, we apologized more for presence-related reasons over IM, probably because we didn't expect availability when we were communicating via SMS (because one or both of us was probably out and about instead of at our computer). Likewise, we probably had more conversations requiring fine-tuned communication over IM instead of text, since long-winded text conversations are a pain. Finally, sympathy-event apologies were more likely to occur over text, probably because life things tend to happen more often when you're out living life, rather than sitting behind your computer.

Okay, deep breath. We've just reached the point where I got kind of insane with the data and further broke down each category into sub-categories. Hover, click, or tap on any of the sub-categories if you want to see a definition.



After hours of comparative analysis, I realized that it's not very valuable to compare our apology behavior to one another (at least, not without a lot more data). For one thing, Ad apologized more than me in just about every category, and even proportional comparisons were misleading. For another thing, there are too many ways to interpret things, and comparing yourself to other individuals is kinda silly anyway. In fact, there was only one comparison that I thought was insightful:

PROPORTION OF JOKING APOLOGIES IN ALL CAPS

Clearly I was the more enthusiastic joker.

Anyway, it's way more illuminating to break down apologies by person.

Ad's Apologies

Robin's Apologies

Ad and I had the same top 3 categories, which I think is interesting. He tended to apologize most for availability-related apologies, and my apology types were more varied.

There are many pictures you can paint with this sort of information. As illuminating as this tracking project was, it was very specific to my dynamic with a particular significant other. I can't help but wonder what my apologies looked like with other people I dated (particularly in non-distance settings), how verbal apologies compare, how my friend/family/stranger apologies align, etc etc etc. Just imagine if I'd tracked all of my apologies to everyone...