4 ms·
The post you're responding to this qualified it with "as long as both participants are willing to yield to interrupt-denies and counter-interrupts when interrup
by awillen 4y ago
The post you're responding to this qualified it with "as long as both participants are willing to yield to interrupt-denies and counter-interrupts when interrupting" - that addresses your point. This doesn't necessarily work well in a context where people are unfamiliar, but with close friends/colleagues/family who are all on board with this conversation paradigm and familiar with each other's interruptive style, it works very well.
- taeric 4y agoIt really doesn't, though. For one, it is recursive. Can I interrupt a counter? What if it is already in a counter? Does it depend on why I'm interrupting? That is, this all only works if you know what "plays" each side can do, and you agree on when each one can be played. Card games with instants captures this remarkably well. In that you have a finite number of "interrupt" cards, and are often limited to how many times you can play each one.
- westoncb 4y agoThe rules (although I'll argue none are needed, see below) really are quite simple: yes, what I described above works recursively: there is no difference between a counter-interrupt and a counter-counter-interrupt. All of this hinges on and ends up being managed by each person giving the other a certain amount of benefit of the doubt, that any of these interrupt-related actions are being done out of a genuine desire to communicate effectively (vs someone trying to make some kind of power play). With that trust in place you don't need to know any "rules", this all emerges quite naturally.
- taeric 4y agoMy assertion is that it is very very easy to miss "easy rules that were emergent and are working well" with "I didn't realize there is a social hierarchy here that I'm massively benefiting from." Or, worse, "there could be consequences for my interrupting this person, I should keep quiet and let it go."
- awillen 4y agoBut again, this isn't going to be the case among people who know each other well. My family and I are interrupters, and we've been at it for decades now - none of the issues you're raising would apply. In terms of knowing when/whether you can interrupt the interrupter/counter-interrupt/etc., see my other post in this thread about how the time of interrupters talking over each other is a signal to both that's constantly evolving (because the increasing time is an increasingly strong signal).
- taeric 4y agoFor certain. I'm less arguing that it isn't a conversation style, than I am that it is a dangerous one. Even in families, it is easy to see this form lead to resentment between folks. Or a contest to be last person talking. In a large sense, mansplaining is similar. Echoing back and explaining things even to people that know them isn't, itself, bad. But it is very easy to portray in a negative light and should not be the assumed default. Now, I fully grant that the largest poison in all of this is the transactional view of conversation. The interrupt style is still predicated on an open transaction that will eventually commit.
- trashtester 4y ago> My assertion is that it is very very easy to miss "easy rules that were emergent and are working well" You are right, at least for people who have poor social skills, are emotionally unstable or immature in some way, including people with autism, ADHD, dementia, children, etc. With such people arround, it may be better to have clearly defined rules for who speaks when. The same goes in situations where people come from very different background or there are very low levels of trust to start with. The better a group of people are at communicating in a cooperative manner, the fewer hard rules are needed.
- xani__ 4y ago
- westoncb 4y agoThat's a good addition: if the participants do not trust one another, the interrupt-denies and counter-interrupts are not going to work. As far the actual signaling goes, AFAICT it's pretty built-in (though I can imagine in other countries for instance there are likely enough differences that you'd have to take time to learn specific signals): I can go into this style of conversation immediately with strangers as long as we're at least giving each other a little benefit of the doubt.
- taeric 4y agoRole play a policemen and a minority in this form of communication to see an easy way it can fall flat. Disastrously.
- westoncb 4y agoIt's already been said plenty of times that mutual trust is a requirement, this will be my last reply to your comments if you keep straw-manning.
- taeric 4y agoIt isn't a straw man, though. It literally happens all the time. Role play a woman in an office. And, I should have stated flat out I can see this style can exist. My discourse is why not to lead in with it. Pretty much anywhere.
- westoncb 4y agoI think I see what you're saying, and it makes sense. But in a scenario with abusive power dynamics injustice is going to play out the same regardless of preferred conversation style, an outlet will trivially be found.
- taeric 4y agoOk, I grant that going full on abusive is not helpful. Such that I apologize on jumping there immediately. :( My concern is more in the ones that aren't quite as stark. Role play that you are talking to someone that doesn't feel safe enough to tell you that they don't feel perfectly safe. Now, how do you know this? If they felt safe, you could ask. But, we have established that they don't, per the role play. And this is a large part of why the conversation around "privilege" is so tough. At large, what we call "privilege" can be easily recast to more subtle dynamics of folks that are allowed to give voice without as much interruptions. Note, not none. Just fewer. (Note that there are obviously other aspects to this that don't necessarily make the recasting.)