I don't know what is the "scientific" term for that but I noticed that, when people give negative feedback, they sometimes use "tricks" to make their words less harsh.

For example, they will use smileys or exclamative ("!", "?") or sentences like: "I might be wrong, but...", "It seems to me that...", etc..

I call those "language smoother" and I remember reading an article about that saying that women use those "tricks" far more than men.

(Edit: here is the article I'm talking about and here is the study to support it)

What I would like to know is: How do "language smoother" work (from a psychological point of view) and how to use them properly (to avoid being seen as childish, unprofessional, or lacking self-confidence)?

And, as an aside, if there is a proper term for that, I would be glad to know how exactly it is called?

Note that I'm only interested in written communication and mostly when it comes to interacting with people you don't know or don't know well.

I'm highly interested in external sources, especially scientific studies on the subject but I will still accept personal experiences as a backup.

  • Related (external website)
    – CaldeiraG
    Commented Oct 4, 2019 at 11:08
  • 1
    Hello network visitors! Please note that IPS is fairly strict about using comments as intended. Comments are only for clarifying and improving the question. Partial answers or general thoughts about the situation may be deleted without notice. If you'd like to write an answer, make sure to check out our posts on How do I write a good answer? and citation expectations first. Thanks!
    – Em C
    Commented Oct 5, 2019 at 16:55

3 Answers 3


The technique you are referring to is called escalation limiting, or more informally, disarming.

disarming, adjective

  • allaying criticism or hostility

Receiving feedback often makes people uncomfortable, which can in turn cause them to become defensive or even hostile to the person giving the feedback. I've been on both sides of this coin, and when the receiver is not receptive, things get really bad.

There are a few different things in play here so I'll go through them one by one.

Claiming non-expertise

Statements such as "I may be wrong, but" are the disarming technique that I've personally found to be the most helpful. The kinds of statements present the feedback to someone in a way that feels less like a personal attack to the person on the receiving end. Rather than asserting something as a fact, you present it as an option, and let the recipient come to their own conclusion. A common scenario where I use this technique is giving feedback to my coworkers about their estimations. Let's take a look at two ways I could give this feedback and then I'll talk about my results.

  1. Your estimations are a bit off. You estimated 2 days for this feature, but it actually took you a week.

  2. I may be wrong, but it seems like your features are taking a bit longer to complete than your estimates.

The first option is very cold and factual. I've gotten feedback like that, and it always feels like the person is saying "You are doing bad and it's not acceptable". When I've gotten feedback using the second method, it ends up feeling almost like the person is asking if I've noticed the problem. When I've given feedback using the second method, it has consistently opened up healthy dialogue where we've been able to talk about the problem and work on a solution (there was one exception to this where I had a coworker who was completely unreceptive to any feedback, but those cases are rare).

I statements

The other major thing happening here is "I statements".

I-statements, also called "three-part messages" or "I-messages," are a way of communicating a problem to another person without accusing them of being the cause of the problem. The formula for these messages takes this form: "I feel (emotion/s) when (circumstance/s)." One can also include the effect that the event has on you.

The point of the I statement is to focus the conversation on what the result of the person's actions was, rather than on their actions themselves. By focusing on the result, you allow them to make the connection between their actions and the outcome. People are more likely to accept a conclusion that they come to themselves, so providing them this opportunity makes it much easier for the feedback to stick.

  • The phenomenon has also often been referred to as hedging, though that term may be losing popularity.
    – Upper_Case
    Commented Oct 4, 2019 at 14:35
  • I would argue that in your feature estimation example, the first sentence is actually better feedback because it's precise and actionable. I guess you could soften the blow by saying the second sentence first, and then add "For example, with feature X..." (Or just say the second sentence first and then add details if needed depending on the other person's response. That seems like what you're going for in your argument.)
    – Llewellyn
    Commented Oct 5, 2019 at 16:35
  • @Llewellyn It will depend on the person. There are certainly some people who prefer the direct approach, but in my experience, the second approach works more often.
    – Rainbacon
    Commented Oct 5, 2019 at 16:44
  • Option 2 works better and feels less artificial to both parties if you make the effort to consider reasons/perspectives from which you might actually be wrong and incorporate them into your wording. Even if you're almost sure there's negative feedback to be given, the scope of things you don't/can't know about that might be contributing is pretty large. Commented Oct 5, 2019 at 22:35
  • @Upper_Case Actually, to my mind, hedging is when you use such phraseology to avoid being seen as completely wrong in the event that you are mistaken. It's the practice of weakening your assertion to the point that it can never be entirely false. It may look similar, but has an entirely different motivation. Commented Oct 6, 2019 at 2:48

I did an intercultural communication course recently and one of the things that was pointed in the course - quite interestingly - is that your speech patterns and how you construct your discourse is heavily influenced by your mother language, even if you have fluency in a second or third language.

Of course, it also works both ways, since how you perceive input is "tainted" by cultural and social cues from your original culture.

If you use longer phrases built outside of the "standard" order for the language, it seems indirect, and is perceived as "smoother". If you use shorter phrases in the "standard" order, you can be considered "blunt" or even "rude".

(I don't have the formal academic sources on hand, I need to unshelf the course material and look through it, I'll come back to it later)

One example that I have quickly on hand from personal experience is the germanics vs latins.

Context: I'm Brazilian (therefore, latin) and I'm currently living in the Netherlands (germanic) with a very international circle of close colleagues.

In my home country, I'm considered a very blunt, no-nonsense person. Here, the natives consider me diplomatic and sometimes I get feedback that my speech is too convoluted or flowery.

The reverse is also true, sometimes it takes me a while to parse that the other person is not being rude, that's just how their speech pattern and social cues work.


How this speech pattern works is twofold:

First, any kind of hedging (it seems to me, I think, it appears that lately) is literally a less extreme position than the same sentence without it. Compare:

Your work does not meet our standards


I am worried that lately there have been some days when your work has fallen just a tiny bit short of our standards

That's not a "psychological trick" it's actually a lesser claim than the undecorated statement, which most people hear as:

I am completely sure that every day, always, your work is significantly short of our standards

So because the complaint or negative feedback is being diluted and made milder, people don't react as strongly to it. I am not saying you are 0 out of 10, just that you're sometimes 8/10 and I need you at 9/10 all the time.

Second, the longer and more convoluted wordings give the person more possible "compliant" responses as well as more possible responses. The blunt statement is really hard to respond to. What can you say besides "sorry" or "yes it does meet your standards!" But to the longer statement you could say:

  • I agree, I have had some bad days, I will make sure that doesn't happen again
  • Please, do not worry, I am also aware of these bad days and am working to prevent them
  • Yes, I am a little below par but pretty close to it even on my worst days, and remember some days are still well above par
  • I actually think those standards are unrealistic and that most of the department falls below them occasionally

And so on.

If you have negative feedback to deliver, you need to sit with it for a while yourself and understand what you are trying to communicate. If it is "you are not good at this job, you fail every day, we have started the process of firing you and I am taking you aside to do you the favour of letting you start looking for work now instead of the day you're fired" then you don't want to do any of this hedging and diluting. But if it is "last Thursday you were less successful than you should be, you need more days like Friday" then that is what you say. Not "your work on Thursday was unacceptable." That statement is broad and comprehensive, doesn't contain much a person can act on, and is more upsetting than a more diluted and accurate statement.

This is even more important in written communication. In person, a friendly tone, a smile, body language and so on can all soften a blunt statement like "that report was simply unusable." In writing, work hard on being precise, on not overstating the failure, and of focusing on what you expect to be done about it rather than only registering that something was not ok. Perhaps

that report was incomplete (you did not include the Europe data, there were no references), hard to understand (a few graphs would have helped enormously) and needed a copy edit for the numerous spelling and grammar errors. It did not use our report template and you omitted the meta data all reports must contain. I have asked [someone] to redo it.

If you're not that confident, the graph thing is more of an opinion, there were just a few grammar errors etc, you would dilute this list accordingly. So "I think the European data is important to have included" for example.

Then you can ask if the person needs training, if there were issues with this particular assignment, or other things that try to solve the problem. Or if they've done tons of great reports before,

Your previous reports showed none of these deficiencies. I'm going to continue assigning this kind of work to you and look forward to a report that meets our standards for your next assignment.

Just please don't do the "praise sandwich" where you tell them they are usually great, this time they were terrible, and you're sure they'll be great again. Studies have shown that the minute the "terrible" part stops, the listener decides you didn't mean the praise part. It's awful. Just tell them they were wrong, but don't come across more extreme than you actually mean.

  • 1
    Hey Kate, thanks for the answer! I noticed you talk about studies regarding the feedback sandwich. I know my question wasn't about that, but do you have a link for those studies? I'm really curious about that on a personal level (so, this as nothing to do with our backup policy). Also, I think your answer would be much stronger with external sources. So, if you happen to have those, that would be great!
    – Ael
    Commented Oct 5, 2019 at 9:00
  • 1
    Here's my ten year old blog entry: gregcons.com/KateBlog/… it links to articles that I recall linking to studies of some sort... Commented Oct 5, 2019 at 10:57
  • @Ælis the general feeling I personally have with feedback sandwich is the same feeling I have whenever somebody says "I'm not racist, but..." or "I'm not sexist, but...", and then proceeds to say something that contradicts the first phrase. Probably it comes with the adversative conjunction. Commented Oct 7, 2019 at 9:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.