Claim analyzed

Science

“In human communication, 93% of meaning is conveyed through nonverbal cues such as body language and tone of voice rather than through the words themselves.”

Submitted by Vicky

The conclusion

False
2/10

The 93% figure is not a valid rule for human communication generally. It comes from limited studies about conveying feelings and attitudes when verbal and nonverbal signals conflict, not from all speech or all meaning. Nonverbal cues matter greatly in many contexts, but the claim turns a narrow finding into a universal percentage that the evidence does not support.

Caveats

  • The statistic is commonly misquoted from Mehrabian's research, which was limited to affective messages rather than general communication.
  • The claim conflates relational or emotional interpretation with overall meaning, including information conveyed directly by words.
  • Evidence shows nonverbal cues are important, but their influence varies by context; there is no credible fixed 93% share across all communication.

Sources

Sources used in the analysis

#1
PMC (PubMed Central) 2022-02-15 | Nonverbal Auditory Cues Allow Relationship Quality to be Inferred ...
SUPPORT

We found that, using nonverbal content alone, people are 75–90% as accurate as they are with full audio cues in identifying positive vs negative relationships, and 45–53% as accurate in identifying eight different relationship types. The results broadly support Mehrabian’s claim that a significant amount of information about others’ social relationships is conveyed in the nonverbal component of speech. Early claims were made for the pre-eminence of nonverbal cues (the Mehrabian Conjecture), at least with respect to the communication of affect, with more than 90% of the information in conversation-based exchanges conveyed by nonverbal cues like intonation, phrasing, and facial expressions (Mehrabian, 1972).

#2
PubMed Central 2021-02-02 | Nonverbal Behaviors “Speak” Relational Messages of Dominance, Nervousness, and Trust
NEUTRAL

Usually, the metacommunication is considered the nonverbal signals that accompany the verbal content and serves to clarify, amplify or even contradict the verbal content. Results indicate that perceptions of dominance are associated with a louder voice, more expressive facial behavior, more head movement, and more and longer turns-at-talk.

#3
PubMed Central 2024-07-01 | Incorporating verbal and nonverbal aspects to enhance a model of communication in cancer care: a qualitative study
NEUTRAL

This qualitative study involved interviews with cancer care providers to develop a grounded theory model of communication that incorporates nonverbal and verbal aspects.

#4
Minnesota State University, Mankato - Communication Studies and Theatre Arts Journal 2007-01-01 | Communication is 93% Nonverbal: An Urban Legend Proliferates
REFUTE

A much-repeated estimate in the popular literature is that 93 percent of the meaning in an exchange comes from nonverbal cues, leaving only 7 percent to be carried by the verbal utterance. Unfortunately, this estimate is based on faulty analysis. As Burgoon (1985) contends, the subjects in the main study by Mehrabian were communicating feelings and attitudes, and the statistic applies only to those limited circumstances.

#5
University of Nebraska-Lincoln DigitalCommons The Relative Contribution to Meaning of Verbal and Nonverbal Channels of Communication: A Meta-Analysis
REFUTE

The present study finds significant effects of the type of meaning conveyed, the age of the receiver, and the stimuli type used as an independent variable in the primary study. It seems clear that meaning is the product of both verbal and nonverbal information, apparently with little interaction between these channels.

#6
Temple Law School 2021-05-03 | BRAIN LESSONS: THE SEVEN PERCENT DELUSION
REFUTE

The claim comes from early 1970s research by psychologist Albert Mehrabian. Mehrabian was testing 'the liking of one person to another.' Extrapolating findings in this one context (and, of course, without repeated studies validating this assessment) has no foundation. The experiment used photographs, frozen images of facial expression. Scholarship has also acknowledged the limits of Mehrabian’s findings.

#7
WIFCON Myth-Information: Communication is 93% Nonverbal
REFUTE

Albert Mehrabian completed research in 1967 showing the significance of non-verbal cues under very specific conditions: Total Liking = 7% Verbal Liking + 38% Vocal Liking + 55% Facial Liking. These equations were derived from experiments dealing with communications of feelings and attitudes (i.e., like-dislike). Unless a communicator is talking about their feelings or attitudes, these equations are not applicable, as Mehrabian himself stated.

#8
TalkBank Verbal and Nonverbal Communication: Distinguishing Symbolic, Spontaneous and Discursive Categories
NEUTRAL

An original meta-analysis demonstrates that, like verbal symbolic communication, nonverbal analogic (pantomimic) communication is related to left hemisphere cerebral processing. In contrast, spontaneous communication is related to the right hemisphere.

#9
Watermark Learning Correcting the Mehrabian Myth – With Words Only
REFUTE

What’s become known as the 7%-38%-55% rule only applies to very specific communication circumstances. The application of the rule to communication in general is a misinterpretation of Albert Mehrabian’s research done in the 1960s that persists even with efforts on the part of Mehrabian and others to correct the misunderstanding. His initial research involved senders communicating a word for the receiver to interpret multiple times, with various levels of congruence between the word spoken and the accompanying tone and facial expression, specifically when communicating attitudes or feelings with inconsistent cues.

#10
BusinessBalls Mehrabian's Communication Theory - BusinessBalls
REFUTE

The ‘Mehrabian formula’ (7%/38%/55%) was established in situations where there was incongruence between words and expression. That is, where the words did not match the facial expression: specifically in Mehrabian’s research people tended to believe the expression they saw, not the words were spoken. Mehrabian did not intend the statistic to be used or applied freely to all communications and meaning.

#11
LLM Background Knowledge 1967-01-01 | Mehrabian's Original Studies Context
REFUTE

Albert Mehrabian's 1967 studies, published in the Journal of Personality and Social Psychology, examined inconsistent communications where a single word conveyed liking or disliking, with tone and facial expressions varying. The 7-38-55 percentages apply only to such specific cases of feelings and attitudes with conflicting cues, not to general human communication. Mehrabian has repeatedly clarified on his website and in 'Silent Messages' (1971) that the rule does not apply broadly.

#12
Jane Anderson Ep #5: Debunking the Myth that 93% of Communication is Nonverbal
REFUTE

Ever heard that only 7% of our communication is verbal and 93% is nonverbal? This is one of the widely misused pieces of research, and biggest urban legends in communication, psychology, and human behavior. It's been quoted by thousands of so-called experts, but the formula does not explain how communication works in general.

#13
Nonverbal Group How Much of Communication is Really Nonverbal? An Extensive Breakdown
REFUTE

93% of communication is not nonverbal and the 55% body language, 38% tone, and 7% vocal 'rule' is not at all true. The number was based on 2 studies by Dr. Albert Mehrabian on inconsistencies of messages involving feelings and attitudes, and he never said this was a general rule for all communication.

#14
Evidence-Based Mentoring Silent Signals: New Review Highlights the Importance of Nonverbal Signals for Perceived Responsiveness
NEUTRAL

Nonverbal signals are often interpreted more quickly and can carry more weight than verbal communication. The effectiveness of nonverbal versus verbal cues varies depending on the context. When verbal and nonverbal cues are congruent, they reinforce each other, leading to greater perceived responsiveness.

#15
Marketing Management Journal THE RELATIVE IMPORTANCE OF VERBAL AND NONVERBAL COMMUNICATION IN A SALES SITUATION: AN EXPLORATORY STUDY
SUPPORT

The importance of nonverbal communication in sales has been widely acknowledged over the years. Although the external generalizability of these studies to sales settings might be questionable, it appears the overall consensus among authors in the social sciences is that nonverbal communication has more impact on message interpretation and influence than verbal communication.

#16
People Shift Mehrabian's 7-38-55 Communication Model: It's More Than Words
NEUTRAL

Albert Mehrabian's 7-38-55 Communication model says 7% of communication is words, 38% through tone and voice and 55% through body language. Even Professor Mehrabian called out words of caution around the absolute accuracy of these percentages in his work. He noted that his research focused on the communication of emotions and feeling, and acknowledged that contributing factors will be quite different when different topics are being discussed.

#17
Communication Cache THE IMPORTANCE OF NONVERBAL CUES IN JUDGING RAPPORT
SUPPORT

Taken together with past research, the present findings suggest that nonverbal behavior may be relatively more important than verbal behavior. Perceivers with access to nonverbal, visual information were the most accurate perceivers of dyadic rapport. Adding verbal content to these channels showed no evidence of improving their judgment and may have even worsened it.

#18
University of Texas Permian Basin How Much of Communication Is Nonverbal? Why the Unsaid Matters
NEUTRAL

Albert Mehrabian found that communication is 55% nonverbal, 38% vocal, and 7% words only. This is where the idea that the vast majority of communication is nonverbal originated, but does this really mean that less than 10% of information is conveyed in spoken words? The formula applies specifically to situations of emotional communication with conflicting verbal and nonverbal cues.

#19
YouTube - Communication Expert Video 93% of Communication is Non-Verbal? Not Quite! Debunking the Mehrabian Myth
REFUTE

Nonverbal communication accounts for 93% of all communication—this claim is one of the most misleading statements repeated in books, articles, and trainings. It comes from Mehrabian's study on recognizing incongruencies between verbal and nonverbal cues in communicating attitudes, not general communication.

Full Analysis

Expert review

3 specialized AI experts evaluated the evidence and arguments.

Expert 1 — The Logic Examiner

Focus: Inferential Soundness & Fallacies
False
2/10

The claim asserts a universal rule—that 93% of meaning in ALL human communication is conveyed nonverbally—but the evidence chain fatally undermines this generalization. Sources 4, 7, 9, 10, 11, 12, 13, and 16 converge to show that Mehrabian's original studies examined only a narrow, specific condition (communicating feelings/attitudes with conflicting verbal and nonverbal cues), and Mehrabian himself explicitly stated the percentages do not apply broadly; the proponent's rebuttal attempts to reframe the claim as being about 'affective meaning' specifically, but the claim as stated makes no such qualification and asserts a universal percentage for 'human communication' generally. The opponent correctly identifies the equivocation fallacy in the proponent's argument—using evidence about accuracy in judging relationship valence or rapport in specific contexts to validate a sweeping universal percentage claim—and Source 5's meta-analysis directly refutes the claim by establishing that meaning is the product of both verbal and nonverbal channels, making the 93% figure logically unsupported as a general rule. The claim is therefore false as stated, relying on a hasty generalization from highly constrained experimental conditions to all human communication.

Logical fallacies

Hasty generalization: Extrapolating a percentage derived from a narrow experimental condition (communicating feelings/attitudes with conflicting cues) to all human communication in generalEquivocation: The proponent conflates 'accuracy in judging relationship valence in affective contexts' with 'meaning conveyed in all human communication,' treating these as equivalent when they are notFalse equivalence: Treating evidence that nonverbal cues are important or dominant in specific relational/affective contexts as equivalent to proving a universal 93% rule across all communication typesAppeal to authority (misapplied): Citing Mehrabian's research as support for a claim that Mehrabian himself explicitly stated his research does not support
Confidence: 9/10

Expert 2 — The Context Analyst

Focus: Completeness & Framing
False
2/10

The claim omits the crucial limitation that the famous 7/38/55 (and thus “93%”) figure comes from Mehrabian's narrow experiments about communicating feelings/attitudes under incongruent cues, not from studies of general human communication or “meaning” across topics, and multiple sources explicitly flag this as an urban legend/misapplication (Sources 4, 6, 7, 11, 18). With full context restored, evidence that nonverbal cues can be highly informative for certain relational/affective judgments (e.g., Source 1) does not justify a universal, quantified statement that 93% of meaning is nonverbal, so the overall impression is false.

Missing context

Mehrabian's percentages were derived from specific lab tasks about liking/disliking and inconsistent verbal vs facial/vocal cues, not general communication (Sources 4, 7, 11, 18).The claim wrongly generalizes a context-dependent finding about affect/relational inference to all “meaning” in human communication, including propositional/informational content (Sources 4, 5, 6).Even research supportive of nonverbal importance typically shows variability by task and does not validate a fixed 93% share across settings (Sources 1, 14, 16).
Confidence: 8/10

Expert 3 — The Source Auditor

Focus: Source Reliability & Independence
False
2/10

The most reliable sources in the pool are the peer‑reviewed/archival academic items on PMC (Sources 1–3) and the university-hosted critiques/meta-analysis (Sources 4–5); taken together, they do not validate a general “93% of meaning” rule, and Source 1 explicitly treats “more than 90%” as an early Mehrabian-era claim limited to affect/relational inference rather than establishing a universal percentage for human communication. Based on these higher-authority, largely independent sources—especially Source 4 (MSU Mankato) and Source 5 (UNL meta-analysis) explaining the statistic's narrow original context and lack of generalizability—the claim as stated (broadly about “human communication” meaning) is false.

Weakest sources

Source 11 (LLM Background Knowledge) is not an independent, citable primary/secondary source and should be discounted versus peer-reviewed or institutionally published references.Source 7 (WIFCON) is a forum/blog-style post with unclear editorial standards and is weaker than peer-reviewed or university publications.Source 9 (Watermark Learning) is a commercial training/blog source with potential marketing incentives and limited methodological transparency.Source 10 (BusinessBalls) is a business self-help site rather than a scholarly or journalistic authority.Source 12 (Jane Anderson) is a personal site/podcast page and not a rigorous, independently vetted source.Source 13 (Nonverbal Group) is a consultancy/interest-group webpage with potential conflicts of interest and no clear peer-review.Source 19 (YouTube video) is not a primary scholarly source and depends on the presenter's credibility rather than formal review.
Confidence: 7/10

Expert summary

See the full panel summary

Create a free account to read the complete analysis.

Sign up free
The claim is
False
2/10
Confidence: 8/10 Unanimous

The arguments

Two AI advocates debated this claim using the research gathered.

Argument for

P
Proponent Argues TRUE

Peer-reviewed evidence shows that nonverbal channels alone can carry the bulk of socially relevant meaning: listeners using only nonverbal auditory content were 75–90% as accurate as with full audio in judging relationship valence, and the authors note these results broadly support Mehrabian's “more than 90%” nonverbal-conveyed information claim for affective, conversation-based exchanges (Source 1, PMC: “Nonverbal Auditory Cues Allow Relationship Quality to be Inferred…”). Consistent with this, research on relational messaging and influence finds that vocal intensity and expressive facial/head behavior systematically drive perceptions like dominance and trust (Source 2, PubMed Central) and that nonverbal behavior can be more diagnostic than verbal content for judging rapport (Source 17, Communication Cache), supporting the motion's core assertion that meaning is predominantly conveyed via tone/body language rather than words.

O
Opponent Rebuttal

The Proponent commits a classic fallacy of equivocation by conflating accuracy in judging relationship valence or rapport — a narrow affective domain — with the sweeping claim that 93% of meaning in all human communication is nonverbal; Source 1 itself explicitly frames its findings within Mehrabian's original affective context, which Sources 4, 7, and 11 confirm applies only to feelings and attitudes communicated with conflicting cues, not general communication. Furthermore, the Proponent's reliance on Sources 2 and 17 to support the 93% figure is a non sequitur, as demonstrating that nonverbal cues influence perceptions of dominance or rapport in specific relational contexts does not validate a universal percentage claim that Source 5 directly refutes by establishing that meaning is the product of both verbal and nonverbal channels across communication broadly.

Argument against

O
Opponent Argues FALSE

The 93% claim is a well-documented misapplication of Mehrabian's research, as Sources 4, 7, and 11 all confirm that his original 1967 studies examined only the very specific circumstance of communicating feelings and attitudes with conflicting verbal and nonverbal cues — not general human communication — and Mehrabian himself repeatedly clarified this limitation. Furthermore, Sources 5 and 6 reinforce that meaning in communication is the product of both verbal and nonverbal channels, and that extrapolating a narrow experimental finding about emotional incongruence to all human communication has no scientific foundation.

P
Proponent Rebuttal

The Opponent's argument commits a scope fallacy by treating the motion as a claim about all propositional content, while the motion's own wording (“meaning” conveyed via tone/body language) is squarely aligned with affective and relational meaning—precisely the domain where contemporary peer‑reviewed evidence finds nonverbal-only cues preserve most of the relevant signal and explicitly links this pattern to “more than 90%” nonverbal conveyance in conversation-based affect (Source 1, PMC: “Nonverbal Auditory Cues Allow Relationship Quality to be Inferred…”). Moreover, citing general cautions that “meaning is the product of both” (Source 5, University of Nebraska-Lincoln meta-analysis; Source 6, Temple Law School) does not rebut predominance in key interpersonal judgments, and it ignores convergent findings that nonverbal behavior systematically drives relational inferences like dominance/trust and can outperform verbal content in rapport judgments (Source 2, PubMed Central; Source 17, Communication Cache).

Your annotation will be visible after submission.

Embed this verification

Every embed carries schema.org ClaimReview microdata — recognized by Google and AI crawlers.

False · Lenz Score 2/10 Lenz
“In human communication, 93% of meaning is conveyed through nonverbal cues such as body language and tone of voice rather than through the words themselves.”
19 sources · 3-panel audit · Verified May 2026
See full report on Lenz →