OK, I’m steaming.  The white smoke of expertise is pouring out of my ears.  What is it about Albert Mehrabian’s famous study on the ‘silent messages’ we get from non-verbal communications that people get so wrong all the time?  It’s constantly taken by ‘communications professionals’ to mean that ’93 % of what we communicate is non-verbal – so it doesn’t matter what you say, it’s how you look – or how you say it’ or something equally absurd.

Now, there’s a new wrinkle:  http://tinyurl.com/klgd6t.  A very nice communications company has misinterpreted the Mehrabian study in a new way!  Aaaaaargh! 

To be sure, the good folks from Creativity Works start out well in this 3-minute video.  They begin by taking all those communications professionals to task, and rightly so.  Mehrabian never said that 93 % of what we say is non-verbal, so words don’t matter (much).  So far so good.

But then they get carried away putting words back on top of the communications heap.  The narrator announces, ‘It’s just not true that delivery can make or break a presentation.’  If you get the words right, Creativity Works argues, the delivery doesn’t matter. 

I’m afraid this is simply not true.  It misrepresents Mehrabian in a new way, and it’s simply, obviously, and demonstrably wrong.  How many times have you watched a presentation where something non-verbal got in the way of the words?  Perhaps the speaker paced aimlessly but doggedly around the stage until you wanted to scream ‘stop!’?  Perhaps the speaker spoke in endless questions, never making a declarative statement in 60 minutes?  Perhaps the speaker simply droned on in an irritating, nasal voice, never making eye contact with the audience until you spaced out, went away on the wings of thought, and never, ever came back? 

We’ve all had experiences like this, countless times.  Of course, delivery can make or break a presentation.  It happens all the time.  And you know it to be true.

OK, let’s get this done.  Here’s what the Mehrabian study did – and didn’t do.  I’ve actually read the original study – something Creativity Works apparently failed to do.  (I suspect the company only checked out the Mehrabian web site, something CW refers to in the video.)  Mehrabian had an experimenter read words to an audience of college students, single words like ‘love’, in different tones and with different expressions.   Then, he asked the audience how it knew what the speaker really meant.  Where did the audience get the clues for the real intent behind the words?  He wasn’t asking about the words at all, but rather the speaker’s intent.  When asked that, the audience responded that it decoded the intent behind the speaker’s words from visual clues 55 % of the time, and from tone of voice 38 % of the time.  Only 7 % of the time did the audience go to the actual words. 

What was the point of all this?  Mehrabian’s work was all about what he called the ‘silent messages’ – how people communicate implicitly their emotions and attitudes.  His big insight – wait for it – was that when words and non-verbal messages were in conflict, people believe the non-verbal every time.  As the CW people show themselves, for example, through a clever cartoon, when a spouse asks, “Are you still angry with me?” and the injured party responds “No,” with folded arms and an angry tone, only an idiot doesn’t realize that in this case ‘no’ means ‘yes’. 

That’s it.  That’s the Mehrabian message:  we get most of our clues of the emotional intent behind people’s words from non-verbal sources.  And when the two are in conflict, we believe the non-verbal every time.  That’s what Mehrabian said, some 40 years ago, and it still is true and powerful today. 



  1. I have always had a problem with this “93% of what we communicate is non-verbal”. Now you have helped me with most of my disquiet, which began with the presentation of this statistic — for a start, 93% is a curiously accurate figure, which makes it immediately suspect (“spurious accuracy” in actuarial-speak) — but moving on, one then wonders how books manage to get anything across to us at all, and why do blind people manage to grasp quite complex subjects without the benefit of our waving arms and happy feet…
    Statistics which are vaguely represented and interpreted are almost always misleading (and downright wrong).
    End of rant.

  2. Thanks for the comment, Piet. And you’re quite right about spurious accuracy. Mehrabian himself noted that the study was a small one and other, similar, follow up studies have founded numbers averaging 60 – 65 % (or roughly 2/3rds) of clues to decode the emotional intent of speakers (again, not their words, but their emotional intent behind the words — Mehrabian assumed people knew what the words meant). So think of it as a range or roughly 2/3rds and you’re better off.

  3. Hi Nick
    There are two separate issues here:
    1. Opinions on whether delivery can make or break a presentation.
    2. The interpretation of Mehrabian.
    Mehrabian’s 1967 study, properly interpreted, does not have any bearing on whether delivery can make or break a presentation. You believe that delivery can make or break a presentation, Creativity Workz don’t. Both of these are opinions which are neither justified or contradicted by Mehrabian’s research.
    Regarding the interpretation of Mehrabian, I don’t see any evidence from the CreativityWorkz video that they didn’t read the original research. In fact the cartoon that they include and that you describe above does capture the nub of Mehrabian’s findings – as you acknowledge.
    Your description of Mehrabian’s research is not quite accurate. For example,
    Mehrabian never asked the listeners how they knew what the speaker really meant. He simply asked them to rate the attitude of the speaker. He asked some to rate on the basis of tone of voice, some on facial expression, and some on the word alone (note: this is a slight simplification, as there were two experiments involved – but I think it conveys the substance).
    And the audience didn’t respond that they decoded the intent from visual clues 55% of the time etc. Mehrabian used statistics to derive those figures.

  4. I’d like to respond to your comments on our video. Our purpose in making it was to draw attention to the misapplication of ‘Mehrabian’s rule’ to all spoken communication – and not to deny that non-verbal communication plays an important role. In fact, we make exactly the same points as you do, and we attempt to explain what Mehrabian’s research was really about, ie what happens when verbal and non-verbal elements in a communication appear to contradict each other.
    Of course, in such a short video we cannot go into too much detail about Mehrabian’s work (which I assure you we have read in the original), but we do show that he was not claiming that his ‘formula’ applies to all spoken communication, only to those where such inconsistencies occur.
    The problem is not so much with Mehrabian’s findings, as with the way they have been unthinkingly applied across the board to all spoken communications. This has led to a widespread belief – all too frequently trotted out on presentation skills courses and the like – that words are the least important (only 7%) part of any spoken message. This is patently ridiculous, since without the words, in most cases there would be no message in the first place. Try explaining Mehrabian’s study using just non-verbal communication!
    So what we were aiming to do with our video was to redress the balance in favour of words – to give words the respect they deserve. We believe that presenters should be given the tools and techniques they need to choose and use their words wisely – not misled into thinking that these things don’t matter.

  5. OK, first of all, it’s wonderful that we’re all arguing about the correct interpretation of Mehrabian’s study. It’s about time that we got it right. And it’s clear that both Olivia and Creativity Workz are on the side of the angels here. We’re all in violent agreement about the main point of Mehrabian, and the way his study has been misinterpreted.
    The irony is that Mehrabian’s work was much broader than the one study that always gets talked about. If you want a better understanding of Mehrabian’s contribution to the understanding of verbal and non-verbal communication, read his book Silent Messages: Implicit Communication of Emotions and Attitudes. It’s available in used bookstores and is well worth the read.
    Second, to Olivia — thanks for your comments. You’re technically correct but miss the larger point. Of course, Mehrabian’s study doesn’t answer the question “can delivery make or break a presentation.” It’s a little study about what happens when the words one says and the emotional intent behind the words are at odds. It’s the way that the study has been taken that both CW and I were reacting to — and about which we largely agree.
    So many communications professionals take the study to mean that “words are only 7 percent of the meaning of what you say,” and that of course is completely wrong.
    You, CW and I are all in agreement about that.
    The misinterpretation really centers on that idea of meaning. Mehrabian was comparing words to the attitude behind them, not the meaning of the words themselves. As you know, and as CW knows. And now, hopefully, everyone else knows.
    When I said “what the speaker really meant” what I meant was the attitude or emotional intent behind the word, which was at variance with the meaning of the word. Like the great example in the CW video where the spouse says, “Are you still angry?” and his partner replies, “No,” but she means “yes.” Here, the speaker really means “yes.”
    So let’s not lose the main point here in getting into a hair-splitting argument over the details of the Mehrabian study. Again, we’re in agreement in the main. Let’s keep that in mind.
    Third, to respond to Martha at CW: thanks for your great comments. The only phrase I was disagreeing with in the video was the “It’s just not true that delivery can make or break a presentation.” Because of course a bad delivery can “break” a presentation.
    Forgive me for an American example, but it’s a famous one. Presidential candidate Richard Nixon was generally agreed to have won his debate with Candidate Kennedy on the merits — as those who listened on radio believed. But those who watched on TV found Nixon’s delivery off-putting, and he ‘lost’ the debate accordingly – and the election.
    I was with you completely until the “not true” comment. Your video is a fun, accurate, and much-needed corrective in all other respects. And I applaud and heartily second your comments above. We are all in agreement here about Mehrabian and the way he’s been misused. When I work with clients I always first work with them to get the words right. The whole point of working on delivery is to get it to reinforce the content, rather than detract from it or undercut it!
    All of us are working to make communications clearer and more effective, and that means primarily getting the message right to begin with!

  6. Thanks for your post, Nick. Like you, I’m delighted to be having this discussion. And yes, I agree that the Nixon interview demonstrates many of Mehrabian’s points very well – it’s a classic example.
    However, I still maintain that the current over-emphasis on the ‘performance’ aspects of a presentation needs to be brought back into balance, and more emphasis placed on the choice and arrangement of the words in the first place. I’m sure you do this as a matter of course, but you might be surprised at the number of presentation skills courses and books that pay words almost no attention all, and usually justify this by citing Mehrabian’s formula, or some version of it. That’s what our video is really about.
    Our video aimed to arouse interest and spark controversy, and it has certainly done that! It is a piece of polemic, and we have used various rhetorical techniques to engage and entertain our audience – including the bold statement that ‘it’s just not true that delivery can make or break a presentation’. In my experience, a well-written speech, badly presented, is generally more effective than the opposite – but perhaps we’ll have to agree to differ on this point!
    If our video gets people thinking about these things for themselves, rather than taking spurious sets of percentages as gospel, it will have served its purpose. So, many thanks for your considered response, and for helping to get the discussion going. As you say, we’re all working towards the same end.

  7. Hi, Martha —
    Good discussion, as you say. Your video is a welcome addition to the ongoing attempt to correct the Mehrabian misunderstanding, which is like kudzu in the American south — it just can’t be eradicated.
    I think of those presentation skills courses and books you refer to as low-level and worthy only to be ignored, but they are out there. It’s astonishing to me how many books on presentations offer a few guidelines about Power Point and suggestions about what to wear and how to stand — and stop there. It is harder work to structure and write a powerful presentation, but if you don’t do that, no amount of clever delivery tricks will save you.
    In the end, what we’re really debating about is which is worse: a well-written speech, badly delivered, or a badly-written speech, well delivered. The short answer to that is that they’re both unfortunate. I’ll give a longer answer in my next blog, and you’re cordially invited to join in…..

  8. The video is really great, well organized and staged. I laughed all the way through.
    Still, any listener will perceive the speaker as a combination of what he says and the way he does that. There’s no doubt, every single detail counts when you aim for audience’s attention. Words, the way we utter them, voice pitch and body position and gestures form a well-played band. Even if one of the instruments plays false, the whole impression will be spoiled for sure.

  9. Hi Nick
    I agree that I was nitpicking – but given that you were setting yourself up as the expert – and claiming that CreativityWorks could not have read the original research – I felt that accuracy was important.
    I also agree that we all agree on the larger point regarding the interpretation of Mehrabian. (BTW I would be interested in references to the follow-up studies you mention).
    On whether delivery can make or break a presentation, here’s my take. I don’t think that delivery can make a presentation – if the content is poor. People may be engaged at the time – but a few hours later they’ll wonder “what was the point of all that”. I call this a “meringue” presentation. Lovely at the time, but it disappears into nothingness. And most of the time, if the content is excellent, I don’t think that delivery can break it – although it may detract. But I do acknowledge that there are instances when the delivery is so bad, that the message does not get across at all. I think that these are rare occasions.
    So, I don’t think the phrase is generally true – it’s far too black and white. I also dislike the phrase because I think it’s not useful to beginner and nervous presenters. Thinking that their delivery is critical to the success of their presentation, tends to make them more nervous and may lead them not to put enough emphasis on their content.

  10. Thanks for the follow-up, Olivia. It is indeed important to be accurate — I was just reading a new book on body language (published in the UK) called “Body Language,” by James Borg, and he not only gets the Mehrabian study wrong, but he also gets the basic brain research wrong!
    As far as follow-ups to Mehrabian, there are many. Lapakko, 1997, Communication Education (46, pp 63-67) has a useful critique, “Three Cheers for Language.” Friedman, 1997, Journal of Experimental Social Psychology, 15, pp 453-469, also finds cases where language matters more in “The interactive effects of facial expression of emotion and verbal messages on perceptions of affective meaning.” Morton & Trehub, 2001, Child Development, 72, pp 834-843 talk about the difference between adult and childhood decoding of affect in “Children’s understanding of emotions in speech.” Further refinements of Mehrabian are researched by Zuckerman and Driver, 1989, Journal of Nonverbal Behavior 13, pp 67-82; by Grahe and Bernieri, 1999, Journal of Nonverbal Behavior, 23, pp 253-269; and by Scherer, 1986, Psychological Bulletin, 99, pp 143-165. K. R. Scherer has many interesting articles on the subject, Speech Communication, 2003; Journal of Cross-Cultural Psychology, 2001; Motivation and Emotion, 1991 — to name only a few.
    But in fact the research has long since passed Mehrabian by, which is yet another reason why it’s ironic that his one study has persisted incorrectly for so long. Look at A General Theory of Love, Lewis, Amini, and Lannon, 2000; The Emotional Brain, Joseph LeDoux, 1996; and Mirrors in the Brain, Rizzolatti and Sinigaglia, 2006, for far more up-to-date views of communication, the brain, affect, and perception.

  11. Hi Nick
    Thank you so much for listing all those references. I have done my own research on this (looking up research papers that cite Mehrabian)but I haven’t read all those that you have listed. So thank you for taking the time to list them out.
    And I agree that the research has greatly moved on since Mehrahbian’s 1967 studies. The longevity of that formula is an amazing example of a sticky idea.

  12. Nick,
    We use some of your videos in our ‘Presentation Skills’ class in my organization! 🙂 Almost every other day, I have arguments about the Mehrabian Myth with my fellow facilitators – I am going to refer them all to this link and not bother to have an argument from now on! Cheers!

    1. Hi, Anita —

      Thanks for your comments. Glad to be helpful in acting as an adjudicator! And glad the videos are useful. It’s time for us all to settle the Mehrabian thing and move on. But most of all, keep up the good work in spreading good presentation habits.

  13. Nick,

    This is just to let you know how much I appreciate this entire discussion. Long ago I despaired of ever seeing a balanced, intelligent discussion of anything on the Internet. Articles were generally narrow and one-sided and the comments were completely polarized, with little or not attention to “where we agree and where we disagree….”

    Yes, I learned a lot and understand the issues much better now. But more importantly, I was very encouraged by the intelligent, thoughtful comments (and rebuttals).

    Good luck, and may all of your work in the future be so fruitful.

    BTW, I’m a communications trainer in Poland, and this information is very helpful….

    1. Hi, John —

      Thanks for the great comment. I agree, too much of the Internet is taken up with mud-slinging matches rather than real exchanges of ideas. Best of luck with your communications business in Poland!

  14. This is a great discussion. I was just brushing up my sketchy memory of Mehrabian through Google when I found it.

    I’ve referred to the percentages in training sessions but only to emphasise that an over-reliance on words – basically reading aloud – loses an audience quicker than almost any other approach. Obviously, if you’re a Shakespeareian trained act

  15. So, for just a fun thought, it has been mentioned that in the scenario with the wife and her husband, when she says no, what she really means is yes due to her body language and tone of voice.

    But, maybe… She means no. I am not a communication expert by any means, but regardless of weather or not she was still up set, she said no and intended that choice of words for a reason. Ultimately, she was lying. And she also may be telling the truth. Perhaps she is mad for a different reason, or maybe she is not mad at him anymore, but simply mad at the situation. I think that sometimes when we assume a person’s meaning based upon their body language, we can easily misinterpret what they want us to know. She may still be upset, but she also may not want anyone to pry any further, so saying no is the easiest way to dismiss further interaction.

    I just think that, though we get a lot out of nonverbal communication, it is just as easily misunderstood.

    1. Thats very clever, and perhaps it precisely takes a free-thinking non-expert to discover such possible exceptions from the obvious. Of course (when youve been inspired to think about it) body”language” though hard (but not impossible) to tell lies with, is very easily misinterpreted. Perhaps many unnecessary conflicts stem from exactly such misreadings of “silent messages”.
      What strikes me as somehow most interesting about the whole matter, is why the manifestly absurd notion of the 3 different meaning-carrying percentages is so extremely popular and beloved despite its logical meaninglessness. This has, I think, not much with the theory as such in it self to do, but with it satisfying some (somewhat strange) needs or longings in the public, perhaps liking to be entertained by freaky facts, and maybe feeling tyrannized by clever words, that emotionally mean a lot less than warm smiles etc. One may experience people getting directly angry when faced with simple objections like: that semantic meaning can not meaningfully be calculated in quantitative percentages; or that results from a single small-scale experiment with a small number of respondents from a narrow selection of socio-cultural backgrounds cannot be meaningfully generalized to general human verbality, despite cultural, linguistic and other known differences. Or for that matter that the good professor himself never drew the mythological conclusions, but reported something quite different. Often such objections will be countered with logically vague, but emotionally forceful explanations about how important the silent messages are. (What is never or seldomly mentioned is the importance of context in interpretation of both meaning and intent in communication. This fact is also interesting, I think.)

Leave a Reply

Your email address will not be published.