Reading Between the Lies: Identifying Concealed and Falsified Emotions in Universal Facial Expressions

Porter and ten Brinke (2008) conducted the first comprehensive frame-by-frame laboratory investigation of genuine versus deceptive facial expressions across five universal emotion categories. The study examined whether emotional leakage — in the form of inconsistent expressions and microexpressions — could be identified in participants instructed to simulate, neutralize, or mask their emotional reactions. Results showed that while true microexpressions were rare, inconsistent emotional leakage occurred in 100% of participants and negative emotions were harder to falsify than happiness.

Adult participants viewed emotionally evocative images and were instructed to respond with genuine or deceptive (simulated, neutralized, or masked) expressions; 697 video clips totaling 104,550 individual frames at 1/30-second resolution were coded for universal expressions, microexpressions, and blink rate, while naïve observer judges rated the veracity of each expression.

Porter and ten Brinke conducted the first comprehensive experimental test of Darwin's hypothesis about involuntary facial expressions revealing deception. Using rigorous frame-by-frame video analysis, they examined three types of deceptive expressions (simulated, neutralized, and masked) across five emotion categories. While emotional leakage was universal among participants, microexpressions were rare and most observers could not reliably detect the deception. The study provided empirical foundation for long-held assumptions about facial communication and deception detection.

Results suggest that while the face does leak genuine emotions during deception, training may be necessary to reliably detect these subtle cues, with implications for investigative interviewing and deception detection practices.

Background & Context

The capacity of the human face to both reveal and conceal emotion has captivated scientists, investigators, and philosophers for centuries. The widespread supposition that aspects of facial communication are uncontrollable and can betray a deceiver's true emotion had, prior to this research, received remarkably little empirical attention. Despite its intuitive appeal and broad cultural acceptance, the idea that the face could serve as an involuntary window into hidden mental states lacked a rigorous experimental foundation.

The intellectual roots of this question trace back to Charles Darwin's landmark 1872 work. Porter and ten Brinke tested a hypothesis originating with Darwin — that there are certain specific facial actions that cannot be created voluntarily, and that facial actions may be involuntarily expressed in the presence of a genuine emotion. This "inhibition hypothesis," long accepted in popular and scientific culture alike, had never been subjected to the kind of fine-grained experimental scrutiny this study provided.

Darwin hypothesized that some facial muscle actions are beyond voluntary control and may "leak" genuine emotion or reveal falsified expressions. Despite uncritical acceptance of this notion in scientific and popular media, little empirical work existed to substantiate these claims prior to this research. Porter and ten Brinke's 2008 paper represented the first comprehensive, frame-by-frame experimental investigation of emotional deception across multiple universal emotion categories, filling a critical void at the intersection of forensic psychology and nonverbal communication science.

Research Design & Methodology

The study was conducted at Stephen Porter's Forensic Psychology Lab at Dalhousie University. Porter and ten Brinke enlisted adult participants to view images that ranged from happy (puppies playing) to fearful (a close-up of an open-mouthed rabid dog) and disgusting (a severed hand), who were instructed to respond with genuine or deceptive emotional expressions. This design allowed the researchers to observe how the face behaves under three distinct deceptive conditions: simulating an emotion not felt, neutralizing a felt emotion (showing a blank face), and masking a felt emotion with a different one.

Key design elements included:

• Five emotion categories examined: disgust, sadness, fear, happiness, and neutral
• Three deceptive expression types: simulated, neutralized, and masked expressions
• 697 total emotion clips recorded and analyzed
• 104,550 individual frames coded at 1/30-second resolution
• Naïve observer judges rated truthfulness without seeing the corresponding images

Each 1/30-second frame — totaling 104,550 frames across 697 expressions — was analyzed for the presence and duration of universal expressions, microexpressions, and blink rate. This painstaking, frame-by-frame coding approach set the study apart from prior deception research, which had typically relied on broader behavioral ratings rather than precise temporal analysis of facial muscle movements. The methodology allowed researchers to identify fleeting expressions invisible to the unaided eye.

Results & Key Findings

The headline finding: emotional leakage is universal but microexpressions are rare. Although untrained observers performed only slightly above chance at detecting deception, inconsistent emotional leakage occurred in 100% of participants at least once and lasted longer than the current definition of a microexpression suggests. This finding was simultaneously reassuring (the face does leak) and sobering (most people cannot detect it).

Key numerical outcomes from the study:

• 100% of participants showed at least one inconsistent emotional expression during deception
• Microexpressions (1/25–1/5 of a second) were exhibited by only 21.95% of participants, appearing in just 2% of all expressions
• Microexpressions were exhibited in the upper or lower face only, not the full face
• Masked emotions were associated with more inconsistent expressions and an elevated blink rate relative to genuine emotions
• Neutralized emotions showed a decreased blink rate
• Negative emotions were more difficult to falsify than happiness

Only 2% of all analyzed videos depicted microexpressions, and almost half of them were shown by truth-tellers rather than liars — a critical qualification that tempers the forensic utility of microexpression detection as a standalone lie-detection tool. The broader category of "inconsistent expressions" (not technically microexpressions) proved far more prevalent and potentially more diagnostically meaningful.

Discussion & Significance

The study's most consequential contribution is its nuanced reframing of what "facial leakage" actually looks like in practice. Popular conceptions, heavily influenced by Paul Ekman's work on microexpressions, had emphasized fleeting full-face expressions as the primary signal of concealed emotion. Porter and ten Brinke's data challenged this, demonstrating that true leakage is more often manifested as inconsistent or out-of-place expressions — partial, asymmetric, or contextually incongruous movements — rather than textbook microexpressions. As ten Brinke noted, "most flashes of inconsistent emotion usually showed in either the upper or lower face only."

The research indicates that negative emotional expressions are more difficult to fabricate than positive ones, suggesting that involuntary leakage of emotion may occur, especially in masked expressions. This asymmetry between positive and negative emotions has theoretical significance: it suggests that the neurobiological circuitry underlying negative affect (disgust, fear, sadness) may be more resistant to voluntary override, a finding with direct implications for high-stakes investigative contexts where subjects may be attempting to conceal genuine distress or fear.

The low detection accuracy among untrained observers underscores a critical lesson. Detecting liars is a tricky business and one that most people — especially people who are highly motivated to catch liars — are particularly bad at. This finding aligns with a broad literature showing that professional investigators, judges, and clinicians perform no better than chance at deception detection without structured tools or systematic facial analysis training. The implication is that the signal is present in the face, but human perception is not calibrated to reliably extract it without guidance.

Limitations & Considerations

Several important methodological constraints limit the direct generalizability of these findings to real-world investigative settings. The study used a laboratory analogue design with controlled image stimuli, meaning that emotional arousal levels — while genuine — were likely lower in magnitude and consequence than those experienced in authentic high-stakes deception scenarios (criminal interrogations, courtroom testimony, etc.). The participants were not personally motivated by fear of punishment or significant reward, which may affect both the intensity of emotional leakage and the sophistication of concealment strategies employed.

• Ecological validity: Lab stimuli (photographs of puppies, severed hands) may not replicate the emotional intensity of real-world deception events
• Observer training: Naïve judges were used; trained coders using FACS or similar systems might yield different detection rates
• Microexpression rarity: The study documented only 14 partial microexpressions across 697 analyzed expressions, indicating they may be too infrequent and unreliable as standalone deception indicators
• Muscle twitches in genuine expressions: Meaningless muscle twitches sometimes occurred even in genuine expressions, meaning correct interpretation can only occur by following up with the right questions
• Sample characteristics: The participant population was drawn from a university setting, limiting demographic diversity

Practical Applications

Porter identified a wide range of potential applications for this research, "from our daily lives to settings like police interrogations, security checks in airports and courtrooms." For forensic practitioners and polygraph examiners, the study reinforces the value of video-recorded interviews that can be reviewed frame-by-frame, and supports integrating systematic facial expression analysis into pre- and post-test interview protocols. The finding that inconsistent emotional displays — not classic microexpressions — are the more reliable and prevalent signal has direct bearing on examiner training curricula.

Crucially, the study cautions against overconfident interpretation of any single facial cue. As ten Brinke observed, "the facial expression appears to crack and another emotion leaks on the face, however briefly" — but this must prompt further inquiry rather than serve as a definitive verdict. For polygraph professionals, this underscores a principle central to best practice: physiological and behavioral signals are probabilistic indicators, not proof, and should always be followed up with structured investigative questioning. The research also points toward a promising future in high-stakes video analysis, with the researchers subsequently collecting over 60 real-world cases of known liars making public appeals — a natural extension into applied forensic contexts.