On the claim that 93% of communication is non-verbal

"The problem with political jokes is they get elected." - Henry Cate VII

***

BBC - BBC Radio 4 Programmes - More or Less, 14/08/2009

"Last week we promised to investigate one of the most-quoted but least-understood statistics of all. That 93% of communication is non-verbal.

Sometimes the claim is more detailed: 7% of communication depends on words, 38% on tone of voice and 55% on body language.

Either way, it's the sort of thing that men in trendy glasses say at management seminars. But, is it true?

These numbers started life in 1968, where they emerged from the work of a Californian psychologist called Albert Mehrabian. And, here we are 41 years later, still talking about them.

So we tracked Dr Mehrabian down, and I started by asking him whether, well, whether 93% of communication is non-verbal.

'Absolutely not, and whenever I hear that misquote or misrepresentation of my findings I cringe because it should be so obvious to anybody who would use any amount of common sense that that's not a correct statement.

If I were to tell you that the pencil you are looking for is upstairs in the desk drawer of the bedroom, three drawers down, I couldn’t do that nonverbally. I mean, I could try to point, but that would hardly locate the pencil, whereas I could do that very precisely with words...

The very first stage of the study was to select a number of words that would clearly communicate positive feeling by themselves. In other words, if the word is written down, the positive words were honey, thanks, *something*, great, love. And then neutral words were maybe, really, so, what. And then the negative words were don't, brute, terrible, no, scram.

Then we had speakers say those words attempting to communicate liking, attempting to communicate a neutral feeling and finally trying to communicate a negative feeling'

'Somebody might say Brute but affectionately'...

'It turned out that when there was a contradiction between the words and the vocal elements, the vocal elements prevailed primarily and very strongly, and the words had hardly any impact. The variance was 5.4 times larger for the vocal elements than for the words.

In a subsequent study I compared the vocal elements with facial expressions. Now in that study, I was able to show very clearly that the facial expressions were 1.5 times more powerful than the vocal elements, and I expressed this as an equation.

Combining the results of the first and the second studies, I was able to arrive at the linear components for content, vocal expression and facial expression, and that's the formula that people have been quoting.'

'Because from your research, you know that the vocal element is 5.5 times more important than the word, so that's 7% vs 38% - about 5.5 times, and we know that the facial expression is 1.5 times more important than the vocal element, so that's the 38% vs the 55%, used to scale it all up.

You were asking a very focused question about the expression of an emotion.'

'Absolutely. There's just no question that you cannot extrapolate my findings to communication in general...

'People keep focusing on this one study, or this pair of studies, and then they keep getting it wrong. Isn't that frustrating?'

'When I look back at the body of work I have produced, I feel I have done much much important work in many areas of psychological research'

'The problem with this, though, is this research which you're more proud of is more detailed, more sophisticated cannot be summarised in an asinine and totally mistaken statistic like '93% of communication is non-verbal'