The Facial Feedback Hypothesis (Pen in Mouth) – Fritz Strack

The philosophical and empirical inquiry into whether our bodily manifestations dictate our emotional reality, or whether internal subjective experience precedes physical display, represents one of the most foundational debates in the history of psychology. For centuries, folk intuition maintained an intuitive, unidirectional chronology: an affective stimulus is perceived, an emotional feeling is generated within the psyche, and this internal state subsequently commands the somatic apparatus to weep, cringe, or smile. However, this classical view was dramatically inverted during the late nineteenth century, inaugurating a sustained scientific inquiry into the bidirectional interplay between peripheral physiology and central nervous system processing. Central to this paradigm is the facial feedback hypothesis, which posits that skeletal muscle feedback from facial expressions plays a causal role in regulating, intensifying, or even initiating subjective emotional experiences.

While theoretical speculations regarding the somatic origins of emotion date back to the foundational treatises of natural history, empirically isolating the specific contributions of facial musculature proved notoriously difficult for over a century. Early experimental attempts were continually compromised by profound methodological artifacts, most notably demand characteristics and cognitive mediation. When researchers explicitly instructed participants to contract their facial muscles into smiles or scowls, participants inevitably recognized the emotional connotations of those expressions. Consequently, observed changes in affective state could easily be dismissed as products of cognitive compliance, self-perception, or social expectancy, rather than genuine, direct physiological feedback operating beneath the threshold of conscious intent. The scientific viability of the facial feedback hypothesis thus stalled at a critical methodological impasse.

This empirical bottleneck was decisively shattered in 1988 when German social psychologist Fritz Strack, alongside his colleagues Leonard L. Martin and Sabine Stepper at the University of Mannheim, published a landmark paper introducing the ingenious “pen-in-mouth” paradigm. By inventing a psychomotor cover story that required participants to hold a writing implement using either their incisors or their lips, the researchers mechanically induced the somatic morphology of a smile or a pout without ever invoking emotional terms, concepts, or conscious motor mimicry. The resulting discovery—that individuals forced into an unobtrusive facial smile evaluated comedic stimuli as significantly funnier than those forced into an inhibiting pout—became an instant classic in cognitive and social psychology. This article provides a comprehensive, exhaustive examination of the facial feedback hypothesis, tracing its intellectual genealogy, dissecting the biomechanical and neuroanatomical realities of the Strack paradigm, charting its central role in the modern replication crisis, and assessing its enduring implications for contemporary embodied cognitive science.

1. Historical Roots and Theoretical Foundations of Facial Feedback

1.1 Darwinian and James-Lange Precursors

The foundational logic underlying the facial feedback hypothesis can be traced directly to the evolutionary inquiries of Charles Darwin. In his seminal 1872 treatise, The Expression of the Emotions in Man and Animals, Darwin argued that facial displays were not merely communicative epiphenomena, but functional adaptations deeply interwoven with emotional survival mechanisms. Crucially, Darwin formulated an explicit early hypothesis regarding somatic feedback: he posited that the free external expression of an emotion by outward signs intensifies it, whereas the repression of all external signs softens or attenuates the emotional state. Darwin noted that even the simulation of an emotion tends to arouse it within our minds, suggesting that physical expressive actions exert an immediate, recurrent regulatory force upon internal feeling states.

A decade later, this peripheralist perspective received its definitive philosophical and physiological formalization through the independent works of American psychologist William James (1884) and Danish physician Carl Lange (1885). The celebrated James-Lange theory of emotion boldly inverted common-sense psychology by asserting that bodily changes follow directly the perception of the exciting fact, and that our feeling of the same changes as they occur is the emotion. In his iconic phrasing, James asserted that humans do not run because they are afraid, or weep because they are sorrowful; rather, we feel afraid because we tremble, and we feel sorrowful because we weep. Under this formulation, subjective emotion is fundamentally the brain’s passive read-out of peripheral physiological shifts, including autonomic arousal, postural alignment, and mimetic facial movements.

This peripheralist model immediately provoked intense academic resistance, culminating in the classic neurophysiological critiques of Walter Cannon and Philip Bard during the 1920s. The Cannon-Bard centralist theory contended that peripheral visceral organs were far too slow, insensitive, and non-differentiated to account for the rapid, nuanced spectrum of human emotional experience. Cannon pointed out that total surgical transection of visceral afferents in animals did not eliminate affective behavior, and that identical visceral patterns characterized entirely different emotional states, such as rage and physiological chill. However, while Cannon’s critique dealt a severe blow to purely visceral theories of emotion, it left open the possibility that the highly differentiated, rapidly responsive, and densely innervated musculature of the human face could provide the precise somatosensory feedback necessary to modulate or differentiate affective experience.

1.2 Silvan Tomkins and Affect Theory

The mid-twentieth-century resurgence of scientific interest in somatic feedback was largely catalyzed by the theoretical architecture of Silvan Tomkins. In his monumental multi-volume work, Affect Imagery Consciousness (1962, 1963), Tomkins developed an elaborate Affect Theory that positioned the human face as the primary organ of emotion. In direct opposition to the prevailing psychoanalytic and extreme drive-reduction paradigms of his era, Tomkins argued that affects were the primary biological motivating mechanism of human life, operating with far greater urgency and flexibility than basic physiological drives such as hunger or oxygen deprivation. Moreover, Tomkins made the radical conceptual claim that affect is primarily facial behavior and sensory awareness of that behavior.

According to Tomkins, evolutionary history endowed the human central nervous system with innately organized, subcortical affect programs. These hardwired neural programs, situated primarily within evolutionary ancient limbic and brainstem regions, are automatically triggered by specific patterns of stimulus density, novelty, and intensity. Once activated, these programs immediately fire coordinated motor efferences to the superficial facial muscles, vascular beds, and vocal apparatus. Crucially, Tomkins maintained that the subjective experience of emotion does not occur until sensory afferents from the contracting facial skin and musculature loop back to the cerebral cortex. The face acts as an amplifying transducer: it translates internal biochemical and neurological activations into distinct, highly salient bodily events that the conscious mind can register as interest, enjoyment, surprise, distress, shame, disgust, contempt, or fear.

Tomkins’ theoretical formulations served as an intellectual springboard for a subsequent generation of empirical researchers, most notably Carroll Izard and Paul Ekman. Izard’s Differential Emotions Theory expanded Tomkins’ framework, proposing that individual primary emotions correspond to discrete, innate neuromuscular patterns whose cutaneous and proprioceptive feedback patterns dynamically construct subjective feeling. Ekman’s ground-breaking cross-cultural studies in the late 1960s and 1970s empirically demonstrated that certain basic facial expressions were universally produced and recognized across non-literate, isolated human cultures. This empirical proof of universal, biologically anchored facial motor configurations laid the indispensable theoretical groundwork for psychologists to explore whether the deliberate or artificial mechanical manipulation of these exact expressive structures could reliably trigger or amplify corresponding emotional states.

1.3 The Methodological Impasse of Early Feedback Research

As affective psychology entered the 1970s, researchers sought to transition the facial feedback hypothesis from abstract theoretical speculation into rigorous empirical science. The earliest pioneering attempts to experimentally test the hypothesis were conducted by James Laird (1974). In his foundational paradigms, Laird placed surface electrodes on participants’ faces—telling them that electrical activity was being recorded for a study on facial muscle activity—and then gave explicit, step-by-step motor instructions to contract specific muscle groups. For instance, participants were instructed to contract the muscles between their eyebrows and draw their lips downward to produce a frown, or to pull the corners of their mouth backward and upward to induce a smile. Participants then viewed emotionally evocative stimuli, such as photographs or cartoons, and reported their subjective affective states.

While Laird’s initial results appeared to substantiate the feedback hypothesis—participants reported feeling happier when their faces were arranged into smiles and angrier when their faces were arranged into frowns—these early methodologies encountered immediate and devastating criticism. The primary objection centered on the catastrophic vulnerability of deliberate muscle-posing paradigms to demand characteristics and experimenter expectancy effects. When an experimenter painstakingly instructs an intelligent human participant to contract their zygomaticus muscles, retract their lips, and elevate their cheeks, the participant rapidly deduces that the experimenter is asking them to produce a smile. Once this conscious cognitive categorization takes place, the participant inevitably activates deeply ingrained socio-cultural scripts regarding the emotional connotations of smiling. The subsequent elevation in reported happiness could easily represent cognitive compliance with perceived experimental demands rather than an unmediated, physiological feedback effect originating from the facial periphery.

A contentious methodological debate erupted throughout the late 1970s and early 1980s. Critics such as Tourangeau and Ellsworth (1979) designed rigorous experiments to directly test whether posing facial expressions of fear or sorrow influenced emotional ratings during the viewing of emotionally evocative film clips. Their results revealed no significant emotional feedback effects, leading them to argue that facial expressions act solely as communicative social displays rather than biological feedback engines of subjective feeling. Tourangeau and Ellsworth asserted that earlier positive findings were entirely artifacts of demand awareness and cognitive labeling. Researchers found themselves trapped in a methodological impasse: how could one physically manipulate human facial musculature with anatomical precision without simultaneously alerting the participant’s conscious mind to the emotional nature of the posture? It was precisely this seemingly intractable puzzle that set the stage for Fritz Strack’s landmark methodological breakthrough.

2. Fritz Strack and the Genesis of the 1988 Landmark Study

2.1 Academic Context of Social Psychology in the Late 1980s

The intellectual climate of late-1980s social psychology was characterized by an overwhelming hegemony of cognitive information-processing models. Human behavior, judgment, and affect were predominantly conceptualized through the lens of computer metaphors: external stimuli were encoded into abstract propositional representations, stored in associative memory networks, retrieved via spreading activation, and integrated through cold, inferential heuristics. In this cognitive orthodoxy, emotions were primarily viewed as downstream, secondary consequences of conscious or subconscious cognitive appraisals. Influential cognitive appraisal theorists maintained that an individual must first interpret and semantically evaluate an event’s personal relevance before any affective reaction or physiological response could manifest.

However, this computational view of human emotion was beginning to face serious challenges from dissenting paradigms. Most famously, Robert Zajonc ignited a fierce academic debate with his landmark 1980 paper, “Feeling and Thinking: Preferences Need No Inferences.” Zajonc argued that affective processing occurs prior to, and independently of, cognitive operations—that affect is primary, fast, and driven by evolutionarily primitive, non-conscious neurobiological pathways. Concurrently, neuroscientists and somatic psychologists were rediscovering the body as an active participant in human thought, anticipating the emergence of what would later become known as embodied cognition. Researchers began investigating somatic markers, visceral reactions, and motor postures as direct, pre-reflective contributors to human cognitive evaluations and aesthetic judgments.

Operating within this stimulating intellectual environment at the University of Mannheim in West Germany, social psychologist Fritz Strack, working in close collaboration with Leonard L. Martin and Sabine Stepper, recognized that the unresolved controversy surrounding the facial feedback hypothesis represented a vital empirical testing ground for these competing models. If facial muscle contraction could be shown to alter subjective emotional appraisal in a context completely stripped of cognitive awareness, it would deliver decisive empirical validation for Zajonc’s non-cognitive affective pathways and decisively refute the necessity of conscious appraisal models. Strack and his colleagues sought to design an experiment that would definitively disentangle visceral somatic mechanics from higher-order semantic processing.

2.2 Epistemic Objectives and Research Hypotheses

The primary epistemic objective of Fritz Strack and his colleagues was to establish an impermeable experimental boundary between pure physical muscle activation and cognitive emotional knowledge. Strack recognized that previous facial feedback experiments had inadvertently confounded two entirely distinct theoretical mechanisms: physiological feedback and cognitive attribution. When individuals consciously adopt a smile, two processes occur simultaneously: first, afferent neuromuscular signals travel from the face to the central nervous system; second, the individual perceives themselves smiling, consciously retrieves the semantic concept of “happiness,” and infers that they must be feeling happy via cognitive self-perception processes. To validate the facial feedback hypothesis as an authentic, biologically grounded phenomenon, one had to systematically eliminate the latter pathway.

Strack, Martin, and Stepper formulated a clear, falsifiable theoretical hypothesis: the contraction of specific facial muscles mechanically associated with positive emotional expressions (specifically, smiling) will elevate subjective feelings of amusement, whereas the contraction of facial muscles mechanically incompatible with smiling will suppress subjective amusement, even when participants are entirely unaware that they are enacting any emotional display. This hypothesis directly targeted the appraisal process. Rather than evaluating whether smiling could spontaneously generate an intense emotion ex nihilo, the researchers aimed to test whether somatic feedback alters the evaluative, subjective appraisal of an ambiguous, external comedic stimulus.

To execute this test, Strack formulated three directional, comparative predictions regarding how participants would evaluate the funniness of humor stimuli under distinct physical states:

• Hypothesis 1 (Facilitation): Participants whose facial muscles are mechanically constrained into a posture facilitating a smile will evaluate comedic stimuli as significantly more amusing than participants in a neutral baseline state.
• Hypothesis 2 (Inhibition): Participants whose facial muscles are mechanically constrained into a posture that physically antagonizes and prevents a smile will evaluate comedic stimuli as significantly less amusing than participants in a neutral baseline state.
• Hypothesis 3 (Independence from Awareness): These affective shifts will occur without participants exhibiting any conscious awareness of emotional themes, experimenter intent, or subjective affective bias, thereby demonstrating that somatic feedback operates prior to cognitive mediation.

2.3 Overcoming Demand Characteristics via Novel Design

Achieving this level of empirical isolation required a completely unprecedented methodological innovation. In designing their manipulation, Strack and his colleagues identified the absolute parameters that any successful paradigm had to fulfill. First, the experimental manipulation had to contract the zygomaticus major muscle—the primary anatomical engine of the smile—with enough sustained force to simulate natural expressive proprioception. Second, the alternative manipulation had to mechanically contract the orbicularis oris—the circular muscle surrounding the mouth—in a manner that rendered smiling physically impossible. Third, and most crucially, the instructions given to participants could not utilize words like “smile,” “frown,” “pleasant,” “unpleasant,” “mouth,” or “emotion.”

Strack realized that human participants possess an extraordinary capacity to generate spontaneous cognitive attributions whenever they are asked to alter their bodily states. If told that the study investigated “facial muscles,” participants would immediately activate emotional schemata. Therefore, the methodological solution required an elaborate, highly credible psychomotor cover story that shifted the participant’s entire attentional focus toward a challenging, non-emotional motor coordination task. The manipulation needed to appear strictly mechanical, difficult, and utilitarian.

The profound brilliance of Strack’s eventual design lay in its absolute simplicity: using a common, inert physical object—a standard writing pen—as an external biomechanical scaffold. By instructing participants to hold the pen with different anatomical structures of their oral cavity, Strack achieved passive, non-voluntary contraction and inhibition of the exact superficial facial musculature under investigation. Participants would not be instructed to “pose an expression”; they would simply be asked to perform basic tasks (such as drawing lines and rating materials) while holding a pen in a designated physical configuration. The pen served as an invisible puppeteer, pulling the facial architecture into affective alignment while keeping the conscious mind completely blind to the puppetry.

3. The Pen-in-Mouth Experimental Paradigm and Methodology

3.1 Cover Story and Psychomotor Deception Protocols

The landmark 1988 experiment, titled “Inhibiting and Facilitating Conditions of the Human Smile: A Nonobtrusive Test of the Facial Feedback Hypothesis,” was published in the Journal of Personality and Social Psychology. Upon arrival at the University of Mannheim testing laboratory, undergraduate participants were introduced to an intricate psychomotor cover story designed to conceal the experiment’s true objectives entirely. The experiment was framed as an applied biomechanical investigation into physical ergonomics, titled an investigation of “motoric coordination with body parts other than the hands.”

The experimenter informed participants that modern biomedical engineering was increasingly interested in developing educational, occupational, and technological tools for individuals with severe physical disabilities, specifically those who had lost the functional use of their arms and upper extremities. Participants were told that researchers were systematically assessing how well individuals could operate writing tools and perform perceptual-motor tasks using alternative muscle systems, specifically the oral and facial structures. This cover story was exquisitely calibrated: it provided a completely logical, compelling, and utilitarian rationale for holding a writing instrument in one’s mouth, while thoroughly deflecting any suspicion away from affective psychology, facial expressions, or emotional appraisal.

To reinforce the authenticity of this psychomotor deception, participants were seated in individual testing cubicles, preventing them from observing other participants or experiencing interpersonal contagion effects. They were administered a series of preliminary psychomotor tasks before encountering the critical experimental stimuli. These preliminary exercises included using the mouth-held pen to draw a straight line between two points on a sheet of graph paper and placing dots inside a series of small, printed boxes. These initial tasks served a double experimental purpose: they firmly established the motor skill context in the participant’s mind, and they allowed participants to acclimate to the mechanical sensations of the pen hold, thereby reducing acute novelty reactions during the critical evaluative phase.

3.2 Biomechanical Manipulation: Teeth, Lips, and Control Conditions

The core independent variable in Strack, Martin, and Stepper’s experimental architecture was the biomechanical configuration used to hold the pen. The researchers operationalized three distinct conditions, each carefully calibrated to produce specific somatic muscular configurations while participants processed visual comedic material:

1. The “Teeth” Condition (Facilitation of Smile): In this condition, participants were instructed to grasp the barrel of the pen horizontally between their front teeth (incisors), ensuring that the pen extended outward laterally from the sides of their mouth. Crucially, the experimenter gave strict instructions that the participant’s lips must not touch the pen at all; the lips had to remain retracted and separated. This precise mechanical posture forced the modiolus and the corners of the mouth backward and upward toward the cheekbones, directly activating and contracting the zygomaticus major and risorius muscles. Biomechanically, this arrangement forced the lower face into the characteristic morphological configuration of a genuine human smile, entirely devoid of subjective emotional intent.

2. The “Lips” Condition (Inhibition of Smile): In this condition, participants were instructed to hold the very end of the pen pointing forward, holding it exclusively with the front portion of their lips. They were explicitly instructed that their teeth must not touch or bite down on the pen under any circumstances. Holding the pen solely by the lips required participants to contract their orbicularis oris muscle into an anterior, sphincter-like purse. This movement physically drew the mouth corners inward and forward. Biomechanically, contracting the orbicularis oris in this fashion creates complete muscular antagonism against the zygomaticus major, rendering it physically impossible to smile while simultaneously mimicking the somatic properties of a pout, a frown, or a state of stern concentration.

3. The “Non-Dominant Hand” Condition (Baseline Control): To establish an empirical baseline against which the teeth and lips conditions could be evaluated, the researchers instituted a neutral control group. Participants in this condition held the pen using their non-dominant hand. The selection of the non-dominant hand, rather than the dominant hand, was a brilliant methodological decision: performing delicate paper-and-pencil tasks with an uncoordinated, non-dominant hand introduces a comparable degree of cognitive effort, unfamiliarity, and mild motor awkwardness, thereby controlling for the general psychomotor difficulty experienced by participants holding pens in their mouths.

3.3 Stimulus Selection: Gary Larson’s ‘The Far Side’ Cartoons

Selecting the appropriate evaluative stimulus was crucial for the success of Strack’s paradigm. If the stimuli were intrinsically hilarious (evoking overwhelming, uncontrollable laughter), the powerful central emotional processing would instantly override any subtle, peripheral somatic feedback signals (ceiling effect). Conversely, if the materials were completely unfunny or offensive, a floor effect would prevent any positive modulation from emerging. The optimal stimulus required a moderate, nuanced degree of humor—an intellectual, mildly absurd comedic form that required cognitive appraisal and subjective interpretation.

To fulfill these exact parameters, Strack and his team selected four single-panel cartoons from Gary Larson’s renowned comic series, The Far Side. Larson’s surrealistic, intellectually playful humor typically depicts human-like animal behaviors, bizarre social paradoxes, and eccentric scientific scenarios. This form of comedic art relies heavily on cognitive incongruity resolution: the viewer must examine the visual frame, detect the unexpected semantic anomaly, and internally resolve the conceptual contradiction to experience amusement.

Because Larson’s cartoons possess an inherent level of conceptual ambiguity, the viewer’s final subjective assessment of whether a panel is “mildly amusing” or “exceptionally hilarious” is inherently subjective and pliable. The researchers carefully standardized the presentation of these cartoons, ensuring that all text captions were uniformly translated into German while preserving Larson’s dry, deadpan comedic tone. The cartoons were presented sequentially in printed experimental booklets, integrated alongside the motor-skill filler tasks, ensuring that participants experienced the evaluation of the cartoons as merely another component of the overarching psychomotor performance battery.

3.4 Dependent Measures and Self-Report Instrumentation

The primary dependent variable in the experiment was the subjective funniness rating assigned to each cartoon. While maintaining their designated pen-holding posture (teeth, lips, or non-dominant hand), participants viewed the four Gary Larson cartoons one by one. Directly underneath each cartoon panel was a calibrated 10-point semantic differential scale, anchored at 0 (“not at all funny”) and extending to 9 (“extremely funny”). Participants used their mouth-held pen (or non-dominant hand) to indicate their evaluations directly on the paper scale.

Crucially, Strack and his colleagues recognized that differences in cartoon ratings could potentially be confounded by extraneous physical variables. If holding a pen with the lips was vastly more physically exhausting or painful than holding it with the teeth, differences in reported funniness might reflect generalized physical discomfort rather than pure facial feedback. To rigorously control for this alternative hypothesis, the post-stimulus questionnaire incorporated secondary self-report measures assessing task difficulty, physical effort, fatigue, and bodily discomfort on corresponding 10-point scales.

Finally, to confirm the absolute success of the psychomotor deception protocol, the experimental session concluded with an exhaustive, funneled debriefing interview. Participants were systematically questioned using increasingly specific prompts:

1. What did you believe was the true, overarching purpose of the experiment?
2. Did you notice any relationship or connection between the various tasks you performed?
3. Did holding the pen in your mouth in that specific position affect your thoughts, your mood, or your evaluations of the cartoons in any way?

Any participant who displayed the slightest inkling of suspicion regarding the study’s emotional focus, or who hypothesized that the pen posture was intended to simulate a smile or a pout, was immediately identified and excluded from the statistical analysis. This stringent filtering ensured that the final dataset comprised only participants who remained entirely blind to the affective nature of the experiment.

4. Muscular Mechanics: Zygomaticus Major vs. Orbicularis Oris

4.1 Neuroanatomy of Facial Musculature in Affect Display

To fully appreciate the elegance of the pen-in-mouth paradigm, one must examine the underlying neuroanatomy governing human facial expression. The superficial muscles of the human face are uniquely distinct from skeletal muscles throughout the rest of the body. Unlike somatic muscles of the limbs, which attach from bone to bone via tendons to actuate rigid joint articulations, facial mimetic muscles attach primarily from the bony framework of the skull directly into the subcutaneous dermis and the superficial musculoaponeurotic system (SMAS). Consequently, their contraction causes complex, fluid shifting, folding, and pulling of the facial skin, creating the distinctive topography of human emotional displays.

The motor innervation of these mimetic structures is governed exclusively by the seventh cranial nerve (Cranial Nerve VII), also known as the facial nerve. The facial nerve originates within the facial motor nucleus located in the caudal ventrolateral pontine tegmentum of the brainstem. Lower motor neurons exit the brainstem, traverse the temporal bone, and arborize through the parotid gland into five major terminal branches: the temporal, zygomatic, buccal, marginal mandibular, and cervical divisions. These efferent pathways carry motor commands that orchestrate the precise, rapid mechanical contractions of the mouth, eyes, forehead, and cheeks.

Equally critical to the facial feedback hypothesis is the sensory afferent architecture. While motor efferences are carried by Cranial Nerve VII, somatosensory and proprioceptive feedback from the facial skin, mucosa, and deep tissues is primarily conveyed back to the central nervous system via the three sensory branches of the trigeminal nerve (Cranial Nerve V): ophthalmic (V1), maxillary (V2), and mandibular (V3). Muscle spindle receptors, Ruffini endings, and Merkel discs embedded within the facial soft tissue continuously fire afferent proprioceptive volleys back through the trigeminal sensory ganglion into the trigeminal sensory nuclear complex of the brainstem. From here, dense ascending projections relay these somatic signals to the thalamus and onto the primary somatosensory cortex, providing the central nervous system with an instantaneous, high-resolution proprioceptive representation of the face’s current mechanical state.

4.2 Isolating the Zygomaticus Major in the Teeth Condition

The physical mechanics of the “teeth” condition were specifically engineered to isolate and sustain isometric contraction of the zygomaticus major muscle. The zygomaticus major is a bilateral, paired strap muscle that originates from the lateral surface of the zygomatic bone, just anterior to the zygomaticotemporal suture. It courses obliquely downward and forward across the cheek, interweaving its muscle fibers at the corner of the mouth into the modiolus—a dense, fibromuscular convergence point where multiple facial muscles meet—and inserting into the dermis of the oral commissure.

When the zygomaticus major contracts during spontaneous or voluntary smiling, it pulls the oral commissures superolaterally (upward and outward toward the ear and zygomatic arch). In the pen-in-mouth paradigm, placing a pen horizontally between the premolars and incisors while demanding that the lips do not touch the object forces the bilateral oral commissures to retract laterally and superiorly. This anatomical positioning acts as a mechanical analog to a conscious smile. The modiolus is drawn taut, elevating the corners of the mouth and producing the classic lateral displacement of the lower cheek pads.

In human affective neuroscience, smiling is famously categorized into two distinct variants: the voluntary, non-Duchenne smile (involving solely the lower-face zygomaticus major) and the spontaneous, genuine Duchenne smile. The Duchenne smile involves the co-activation of the zygomaticus major alongside the pars lateralis of the orbicularis oculi, the sphincter muscle encircling the eye orbit, which raises the cheeks and creates characteristic wrinkling (“crow’s feet”) at the lateral canthus. The horizontal pen-in-mouth manipulation primarily isolates the zygomaticus major. By elevating the lower cheek mass, the pen posture mechanically stimulates the somatosensory afferents associated with the lower-face smiling apparatus, providing the brainstem and somatosensory cortex with continuous proprioceptive signals of positive facial valence while avoiding the activation of antagonistic depressor muscles, such as the depressor anguli oris.

4.3 Inhibition and Antagonism via the Orbicularis Oris

Conversely, the mechanics of the “lips” condition were designed to activate an entirely distinct, functionally antagonistic muscular network centered upon the orbicularis oris. The orbicularis oris is a complex, multi-layered circular sphincter muscle that completely surrounds the human oral fissure. It does not originate from a single bony attachment; rather, it consists of numerous intrinsic muscular strata alongside extrinsic fibers contributed by surrounding buccinator, levator, and depressor muscles that converge into the lips.

When an individual holds a pen pointing directly forward, supported exclusively by the compressive force of the lips without any assistance from the incisors, the intrinsic fibers of the orbicularis oris must undergo sustained isometric contraction. To grasp the foreign object securely, the marginal and labial portions of the muscle contract circumferentially, drawing the lips together, curling the vermilion borders inward, and projecting the lips anteriorly into a pronounced purse or pouting posture. This anterior traction produces direct biomechanical antagonism against the zygomaticus major. The modiolus is pulled firmly anteriorly and medially toward the facial midline, mechanically locking the oral commissures in place and rendering lateral-superior retraction physically impossible.

Beyond this purely mechanical inhibition of the smiling apparatus, the somatic morphology generated by the lips condition corresponds closely to universal emotional displays of negative affect and defensive withdrawal. Pouting, lip-pursing, and oral tightening are universal somatic markers observed in frustration, stern skepticism, concentration, and infantile distress. By activating the orbicularis oris, the lips manipulation flooded the trigeminal afferent channels with proprioceptive patterns characteristic of emotional inhibition, reserve, and disapproval. Thus, the two conditions established a perfect physiological contrast: one mechanically simulated the somatic pathways of enjoyment, while the other physically locked the face into a state antagonistic to amusement.

5. Key Findings and Empirical Results of the 1988 Experiment

5.1 Quantitative Discrepancies in Humor Ratings

The empirical results of Fritz Strack, Leonard Martin, and Sabine Stepper’s 1988 experiment provided immediate, striking quantitative support for the facial feedback hypothesis. In their primary experiment, the dependent variable was computed as the average funniness score across the series of Gary Larson cartoons on the 0-to-9 semantic differential scale. The data revealed a clear, monotonic progression across the three experimental conditions that perfectly matched the authors’ theoretical predictions.

Participants in the “teeth” condition—whose facial muscles were unobtrusively locked into a configuration simulating a smile—rated the Gary Larson cartoons as markedly funnier than participants in either of the other two groups, reporting an average funniness score of 5.14 (with a standard deviation of 1.12). In contrast, participants assigned to the neutral “non-dominant hand” control group rated the exact same cartoons with a moderate average score of 4.77 (standard deviation 1.25). Most dramatically, participants in the “lips” condition—whose facial musculature was constrained into an orbicularis oris contraction that physically inhibited smiling—evaluated the identical cartoons with a severely depressed average funniness score of 4.32 (standard deviation 1.47).

These findings revealed a symmetric, bi-directional somatic feedback phenomenon. The mechanical activation of the zygomaticus major did not merely induce a slight statistical fluctuation; it actively facilitated and amplified the subjective experience of comedic amusement relative to baseline. Simultaneously, the contraction of the orbicularis oris exerted an active inhibitory influence, suppressing humor evaluations well below the baseline established by the control group. Strack and his colleagues replicated this fundamental pattern across two distinct experimental variations within their original 1988 publication, confirming that the effect was robust across slightly different operational sequences.

5.2 Statistical Significance and Effect Sizes

To evaluate the statistical validity of these empirical discrepancies, Strack, Martin, and Stepper subjected their raw quantitative data to rigorous analysis of variance (ANOVA). In their primary experimental trial, the planned omnibus contrast testing the linear trend from the inhibiting lips condition, through the neutral control condition, to the facilitating teeth condition achieved clear statistical significance, yielding an F-statistic of F(1, 88) = 5.63, p < .02. When conducting direct pairwise contrasts between the two extreme facial postures, the difference in humor ratings between the teeth condition and the lips condition was highly significant, t(59) = 2.45, p < .018.

When translated into modern standardized metrics of effect magnitude, the difference between the teeth and lips conditions in Strack’s 1988 study yielded a standardized mean difference (Cohen’s d) of approximately d = 0.62. In the behavioral and social sciences, an effect size of 0.62 is classified as a moderate-to-large effect, indicating that nearly 73% of the participants in the smiling posture rated the comedic stimuli higher than the average participant in the frowning posture. This represented an extraordinarily powerful somatic effect, particularly given that the manipulation involved nothing more than holding a standard writing implement between one’s teeth or lips for a few minutes.

Critically, the authors demonstrated that these robust statistical discrepancies could not be explained away by extraneous physical confounds. Analysis of the secondary control variables revealed that participants in the teeth condition and the lips condition reported virtually identical levels of task difficulty, physical effort, and subjective bodily strain. The ANOVA performed on the difficulty ratings yielded no statistically significant difference between the two mouth groups (F < 1.0), and entering perceived difficulty as a statistical covariate in an ANCOVA model did not diminish the magnitude or significance of the primary facial feedback effect. The variation in cartoon amusement was unequivocally tied to the specific facial muscle configuration, rather than generalized physical annoyance or fatigue.

5.3 Theoretical Resolution of Cognitive Mediation

The successful execution and dramatic findings of the Strack, Martin, and Stepper (1988) study represented a watershed moment in the history of affective science. For over a decade, critics of the facial feedback hypothesis had argued that any observed shifts in emotion following facial manipulation were merely artifacts of cognitive self-perception, demand compliance, or social priming. The pen-in-mouth paradigm shattered this argument by demonstrating that emotional modulation occurs in the complete absence of conscious emotional intent or categorical awareness.

The post-experimental debriefing interviews confirmed that not a single participant in the analyzed sample had connected their physical posture to the concept of smiling, frowning, or emotional valence. The participants genuinely believed they were participating in an ergonomics study assessing psychomotor skills for the physically disabled. Because the participants possessed no conscious awareness that their facial muscles were arranged into an expressive configuration, they could not have utilized cognitive self-perception processes—such as “I am smiling, therefore I must be happy”—to deduce their affective ratings. The mediation was not semantic; it was direct, mechanical, and somatic.

By decisively circumventing cognitive mediation, Strack’s study provided the first definitive, methodologically unassailable evidence that somatic peripheral states exert an autonomous, bottom-up influence upon higher-order evaluative judgments. It validated Silvan Tomkins’ affect theory, provided empirical backing for Zajonc’s assertion that emotional processing can bypass cognitive appraisal, and established the pen-in-mouth paradigm as the undisputed gold standard in behavioral affective research. For the next three decades, virtually every major textbook in introductory psychology, social psychology, and cognitive neuroscience featured Fritz Strack’s 1988 study as canonical proof of the facial feedback hypothesis.

6. Neurobiological and Cognitive Mechanisms of Facial Feedback

6.1 Proprioceptive Feedback and Cortical Mapping

In the wake of Strack’s behavioral findings, neuroscientists and physiological psychologists sought to uncover the precise neurobiological architecture through which peripheral facial contractions alter cognitive appraisals and emotional states. The prevailing model focuses on ascending proprioceptive feedback loops that project from the facial periphery into high-order somatosensory and limbic processing centers. When the zygomaticus major or orbicularis oris contracts, low-threshold mechanical stretch receptors in the cutaneous tissue and muscle spindles fire continuous action potentials along the sensory afferents of the trigeminal nerve (Cranial Nerve V).

These trigeminal afferents enter the brainstem, synapse within the principal sensory nucleus of the trigeminus, and ascend via the trigeminothalamic tract (trigeminal lemniscus) to terminate in the ventral posteromedial (VPM) nucleus of the thalamus. The thalamus, acting as the brain’s primary sensory routing hub, projects these somatic signals directly to the primary and secondary somatosensory cortices (S1 and S2) located within the postcentral gyrus. Neuroanatomically, the human somatosensory cortex exhibits an extreme evolutionary bias: the cortical homunculus reveals that the face, lips, and oral structures occupy a vastly disproportionate volume of cortical real estate compared to the entire rest of the human body. This massive cortical representation reflects the exquisite sensitivity and communicative importance of the human facial apparatus.

From the primary somatosensory cortex, dense reciprocal neural pathways interface directly with key affective and interoceptive structures, most notably the anterior insular cortex and the amygdala. Modern functional neuroimaging (fMRI) studies have revealed that the anterior insula functions as a master integrative clearinghouse for interoception, synthesizing internal visceral states, bodily sensations, and cutaneous feedback into conscious affective feelings. When the brain registers the distinct somatosensory signature of zygomatic retraction, this cortical pattern triggers bottom-up activation within the ventral striatum and orbitofrontal cortex, priming the central nervous system to evaluate ambiguous incoming stimuli through a positive evaluative lens.

6.2 The Vascular Theory of Emotional Efference

While direct proprioceptive cortical feedback represents the dominant theoretical model, a compelling alternative neurovascular mechanism was proposed by Robert Zajonc and his colleagues in the late 1980s: the Vascular Theory of Emotional Efference (VTEE). Reviving and modernizing a nineteenth-century hypothesis formulated by French physician Israel Waynbaum, Zajonc proposed that facial muscle movements regulate subjective affect by physically altering the temperature of the brain.

The VTEE model is grounded in cerebral hemodynamics. The human brain, while comprising only 2% of total body mass, consumes approximately 20% of the body’s resting metabolic energy, generating tremendous metabolic heat. Proper neural function is extraordinarily sensitive to localized fluctuations in cerebral temperature. Zajonc pointed out that venous blood draining from the facial skin and nasal mucosa flows backward through the angular and ophthalmic veins directly into the cavernous sinus—a venous lake situated at the base of the skull that completely surrounds the internal carotid artery as it delivers arterial blood to the brain and the hypothalamus.

According to the vascular theory, contracting the facial muscles alters the volume and velocity of blood entering this cavernous cooling manifold:

• Smiling Mechanics: Contracting the zygomaticus major while parting the lips increases nasal airflow and compresses specific facial veins, diverting cooled cutaneous blood into the cavernous sinus. This causes localized thermoregulation—a slight drop in hypothalamic temperature—which promotes the release of positive neurotransmitters, such as serotonin and beta-endorphins, inducing subjective feelings of pleasantness.
• Frowning Mechanics: Conversely, contracting the orbicularis oris, corrugator, or procerus muscles constricts nasal passages and restricts cooling blood flow, causing a subtle rise in brain temperature. This localized hyperthermia is hypothesized to trigger neurochemical cascades associated with subjective irritability, distress, and negative affect.

While modern neuroimaging has demonstrated that proprioceptive neural signaling plays a far more immediate and primary role than vascular thermoregulation, the VTEE framework remains an intriguing historical attempt to provide a purely mechanical, non-cognitive physiological explanation for Strack’s experimental findings.

6.3 Dual-Process Models and Embodied Cognition

Beyond neurobiology, the cognitive mechanisms linking Strack’s pen manipulation to elevated humor ratings can be elegantly explained through the framework of dual-process cognitive architectures. In modern cognitive psychology, mental functioning is conceptualized as an ongoing interaction between two distinct operational modes: Type 1 (fast, automatic, non-conscious, heuristic-driven processing) and Type 2 (slow, deliberate, conscious, rule-governed analytic processing). Under this dual-process paradigm, facial feedback functions as a low-level, pre-reflective Type 1 heuristic cue.

When an individual encounters a Gary Larson cartoon, their cognitive system must rapidly process complex semantic incongruities. In evaluating whether the cartoon is “funny,” the individual does not engage in a detached, algorithmic calculation of comedic value; instead, their Type 2 analytical system queries internal subjective feelings: “How does this make me feel?” In psychological science, this is known as the “affect-as-information” heuristic, famously formulated by Norbert Schwarz and Gerald Clore. Rather than inspecting the stimulus in absolute isolation, people consult their immediate, ambient bodily state as diagnostic information regarding the value of the target.

In Strack’s paradigm, participants in the teeth condition already possessed an activated somatic substrate of amusement. The continuous proprioceptive feedback from the contracted zygomaticus major provided the Type 1 system with a faint, ambient somatic signal of pleasure. When the cognitive system evaluated the cartoon, this somatic signal was automatically and misattributively integrated into the evaluative process. The cognitive appraisal system resolved the ambiguity by concluding: “This cartoon is producing pleasant sensations, therefore it must be very funny.” This theoretical synthesis bridges the gap between Fritz Strack’s behavioral findings and the broader architecture of grounded and embodied cognition, demonstrating that abstract human judgments are deeply rooted in, and continuously modulated by, the physical morphology of the living body.

7. Early Critiques, Methodological Scrutiny, and Boundary Conditions

7.1 Ecological Validity and Artifactual Stressors

Despite its immediate canonization, Fritz Strack’s pen-in-mouth paradigm was not entirely immune to methodological skepticism. As the initial excitement surrounding the 1988 paper settled, critics within cognitive and physiological psychology began raising important concerns regarding the ecological validity and potential biomechanical artifacts inherent in using an external foreign object to manipulate the human face. The most fundamental critique centered upon the sheer artificiality of the experimental arrangement: humans do not, in their everyday social lives, navigate the world with rigid cylindrical plastic implements wedged horizontally between their teeth or pursed within their lips.

Physiological critics noted that placing an inorganic foreign object into the human oral cavity immediately stimulates a cascade of involuntary protective and digestive reflexes. Holding a pen horizontally between the incisors stimulates the periodontal mechanoreceptors embedded within the periodontal ligament, sending compressive feedback through the alveolar nerves. Furthermore, foreign oral objects stimulate the salivary glands, causing an immediate, unnatural hypersecretion of saliva (hypersalivation). Participants in both the teeth and lips conditions were frequently forced to swallow repeatedly, fight the urge to drool, or adjust their jaw muscles to prevent muscular cramping of the masseter and temporalis muscles.

These observations led some theorists to question the precise psychological valence of the somatic state induced by the pen. Did the teeth condition truly evoke a pure, authentic somatic facsimile of joyous amusement? Or did holding a pen between the incisors merely create a state of physiological tension, heightened arousal, and dental stress that happened to push ratings upward, while the lips condition produced a state of oral fatigue? Skeptics argued that by introducing an awkward, foreign object into the oral cavity, the paradigm introduced confounding tactile and proprioceptive artifacts that bore little resemblance to the spontaneous, effortless, and holistic motor coordination of a natural, human emotional expression.

7.2 Physiological Fatigue versus Emotional Modulation

A second major line of methodological inquiry focused on the potential confound of asymmetric muscular fatigue. When evaluating Strack’s experimental groups, several researchers postulated that holding a pen horizontally by biting down on it with the incisors relies primarily upon the massive, powerful skeletal muscles of mastication (the masseters, temporalis, and medial pterygoids) supported by the skeletal resistance of the jawbone. Biting down requires relatively little active muscular strain once the jaw is locked into position.

In stark contrast, holding a pen pointing directly outward, suspended entirely by the fleshy, unsupported muscular tissue of the lips, places an immense, continuous isometric burden on the orbicularis oris. The orbicularis oris lacks any rigid skeletal articulation; it is a soft-tissue sphincter. Sustaining an isometric forward pinch against the weight and leverage of a protruding pen rapidly leads to local muscular ischemia, lactic acid accumulation, and fine motor tremors. Consequently, critics proposed a parsimonious, alternative physiological explanation for Strack’s findings: the depressed humor ratings observed in the lips condition did not reflect an authentic emotional “inhibition of smiling,” but rather a simple, mundane state of physical exhaustion, muscular burning, and irritation.

While Strack and his colleagues had anticipated this critique by measuring subjective difficulty and effort in their 1988 post-test questionnaires—reporting no statistically significant differences in perceived difficulty between the teeth and lips groups—subsequent critics argued that subjective self-reports of “difficulty” may not adequately capture localized, subconscious neuromuscular fatigue. If the lips group experienced subtle, unmeasured physical discomfort, any cognitive evaluation performed under that somatic duress would naturally be skewed negatively. This debate highlighted the profound technical difficulty of perfectly calibrating physical exertion across distinct muscle groups when testing embodied psychological phenomena.

7.3 Cross-Cultural and Contextual Boundary Conditions

As researchers globally attempted to build upon the 1988 findings, it became increasingly apparent that the facial feedback effect was subject to powerful, delicate contextual and individual boundary conditions. One significant moderating factor was the nature of cultural display rules and somatic socialization. In cultures where the overt display of emotion is heavily regulated or discouraged (such as certain East Asian societies), individuals develop heightened metacognitive monitoring over their facial expressions. Cross-cultural affective researchers noted that when individuals possess deeply internalized norms against outward emotional display, the subtle bottom-up cues generated by peripheral facial contractions are far more likely to be suppressed or overwritten by top-down regulatory mechanisms.

Furthermore, the physical and social context of the testing environment proved to be an immensely influential, albeit initially underappreciated, boundary condition. In Strack’s original 1988 setup, participants were isolated inside individual cubicles, completely shielded from social observation. Subsequent studies revealed that if an experimenter remained standing directly over the participant, or if testing occurred in a shared communal room, the facial feedback effect frequently vanished. The presence of social observers immediately shifts the psychological function of facial displays from private somatic experiences to public, communicative social signals, fundamentally altering how the central nervous system decodes the meaning of facial muscle tension.

Finally, the modality and baseline intensity of the stimulus material established rigid boundary conditions. When researchers attempted to use the pen-in-mouth paradigm alongside emotionally heavy or unambiguous stimuli—such as graphic video clips depicting severe medical surgeries, deeply tragic literature, or blatantly hilarious stand-up comedy performances—the peripheral feedback from the pen manipulation had zero measurable impact on participant ratings. The facial feedback effect appeared strictly confined to a narrow, delicate evaluative band: it could nudge the needle when an individual evaluated ambiguous, mildly humorous, or emotionally neutral materials, but it was easily obliterated whenever strong, unambiguous top-down emotional or cognitive signals were present.

8. The Replication Crisis and the 2016 Registered Replication Report

8.1 The Wagenmakers et al. (2016) Protocol Architecture

By the early 2010s, the field of social and cognitive psychology was violently shaken by what became universally known as the replication crisis. The discovery of widespread publication bias, dubious research practices (such as p-hacking, selective reporting, and post-hoc hypothesizing), and high-profile failures to replicate classic, foundational psychological experiments forced the scientific community to re-evaluate its empirical literature. In response, leading scientific bodies and journals established rigorous, pre-registered replication protocols designed to evaluate whether psychology’s most celebrated findings could withstand massive, highly powered, transparent empirical scrutiny.

Given its iconic status as the definitive, textbook demonstration of non-cognitive facial feedback, Fritz Strack’s 1988 pen-in-mouth study was formally selected by the Association for Psychological Science (APS) for a prestigious Registered Replication Report (RRR). Coordinated by Dutch psychometrician and methodologist Eric-Jan Wagenmakers, this colossal replication effort was designed to provide the most definitive, statistically unassailable test of the pen-in-mouth effect ever attempted. The protocol, published in Perspectives on Psychological Science in 2016, mobilized an unprecedented international coalition comprising 17 independent laboratories distributed across the globe, spanning North America, Europe, and Asia.

The Wagenmakers et al. (2016) initiative eliminated virtually all the historical vulnerabilities that had plagued earlier psychological science:

• Extreme Statistical Power: While Strack’s original 1988 sample featured 92 participants, the multi-lab 2016 replication amassed a gargantuan sample of 1,894 participants, providing virtually 100% statistical power to detect even tiny effect sizes.
• Pre-Registration: Every detail of the experimental protocol, stimulus presentation, participant exclusion criteria, and statistical analysis pipeline was formally pre-registered and publicly locked prior to the collection of a single data point.
• Standardization: The instructional scripts, participant interactions, and testing materials were completely standardized. All experimental trials were computerized, using uniform software running on identical monitors to display the Gary Larson cartoons and record self-report ratings with millisecond precision.
• Consultation with the Original Author: Fritz Strack was directly consulted during the planning phase. He generously shared his original instructional scripts, provided the exact set of Gary Larson cartoons utilized in 1988, and reviewed the proposed experimental protocol.

8.2 Empirical Outcomes Across the Seventeen Laboratories

The empirical outcomes of the Wagenmakers et al. (2016) Registered Replication Report sent seismic shockwaves throughout the international psychological community. Across the 1,894 tested participants, the classic, foundational facial feedback effect utterly failed to materialize. When the data from all 17 independent international laboratories were synthesized through a random-effects meta-analytic framework, the difference in humor ratings between participants holding the pen in their teeth versus those holding it in their lips was statistically indistinguishable from zero.

The quantitative results were devastatingly unequivocal. The pooled meta-analytic effect size across the 17 participating labs was a microscopic d = 0.03, with an exceptionally narrow 95% confidence interval spanning from -0.11 to +0.16. This confidence interval cleanly encompassed zero, confirming that the observed difference had no statistical significance (p = .51). To make matters worse for the classical hypothesis, not a single one of the 17 individual laboratories succeeded in producing a statistically significant replication of Strack’s original 1988 finding (p < .05). In fact, out of the 17 participating sites, nine laboratories observed tiny positive numerical differences in the predicted direction, while the remaining eight laboratories actually observed numerical differences in the opposite direction (where participants in the lips condition rated the cartoons as funnier than those in the teeth condition).

A comprehensive Bayesian meta-analysis conducted on the pooled dataset revealed a Bayes factor (BF₀₁) strongly favoring the null hypothesis over the alternative hypothesis. The statistical data indicated that the observed results were over 30 times more likely to occur under a model where the pen-in-mouth manipulation exerts zero effect upon humor evaluations than under a model reflecting Strack’s originally reported effect size. The quantitative data did not merely fail to reach statistical significance; it actively provided overwhelming evidence in favor of the null hypothesis.

8.3 Initial Scientific Fallout and Institutional Reactions

The publication of the 2016 Registered Replication Report was treated by science journalists, university faculty, and social psychologists as an empirical earthquake. Prominent media outlets—including The Atlantic, Slate, and The New York Times—ran prominent feature articles announcing that one of the most famous psychological discoveries of the twentieth century had been thoroughly debunked. The headline was simple and sensational: “Smiling Does Not Make You Happy.” Textbook publishers and psychology professors faced immediate pedagogical dilemmas: should the pen-in-mouth study, long taught as an immutable biological reality of human nature, be immediately excised from educational curricula?

Within academia, the RRR results catalyzed a ferocious debate regarding the epistemological integrity of classic social psychology. Critics of the discipline pointed to the 2016 collapse as irrefutable proof that the mid-to-late twentieth-century psychological literature was riddled with false-positive results, generated by underpowered samples, publication bias, and unmonitored researcher degrees of freedom. Methodological reformists cited the Strack failure as an unassailable justification for the mandatory adoption of open-science practices, universal study pre-registration, and large-scale multi-site replication initiatives.

However, amidst the widespread public declarations of the demise of the facial feedback hypothesis, more cautious theorists urged intellectual restraint. They emphasized a crucial epistemological distinction that was frequently lost in sensationalized media accounts: the failure of the Wagenmakers et al. (2016) multisite replication did not mathematically disprove the broad facial feedback hypothesis in its entirety. What it proved was that the specific, operationalized pen-in-mouth paradigm, as executed under the exact parameters of the 2016 computerized protocol, was profoundly unreliable. Affective scientists began asking a critical question: what subtle, unnoticed methodological difference separated the German testing cubicles of 1988 from the standardized computer labs of 2016?

9. Fritz Strack’s Rebuttal and the Video Camera Confound

9.1 Identification of Key Methodological Discrepancies

Fritz Strack did not accept the conclusion that his 1988 finding was a historical illusion. Accompanied by the publication of the 2016 RRR, Strack published a formal, sharp academic rebuttal in Perspectives on Psychological Science, critically dissecting the replication’s methodology. Strack argued that while the 17 replication laboratories had successfully standardized the superficial mechanics of the experiment—using identical software, scripts, and cartoons—they had committed a fatal methodological deviation from his original 1988 experimental environment: the mandatory introduction of video cameras placed directly in front of the participants.

In designing the 2016 replication protocol, the coordinators had faced an operational challenge: how could experimenters ensure that participants in individual cubicles were strictly adhering to the difficult pen-holding instructions (keeping their lips retracted in the teeth condition, and avoiding touching their teeth in the lips condition) without having an experimenter physically intrude into the room? The replication coordinators devised what they considered an elegant, harmless quality-control measure: they mounted a visible video camera directly on top of the computer monitor, pointing straight at the participant’s face, to continuously record their oral posture throughout the task. Participants were explicitly informed that they were being recorded, and video recordings were subsequently reviewed to discard non-compliant participants.

Strack pointed out that this seemingly innocuous procedural addition fundamentally sabotaged the psychological validity of the experiment. In the original 1988 study, there were no video cameras, no digital monitoring devices, and no electronic surveillance. Participants were completely alone in an analog, private cubicle. Strack contended that pointing a video camera directly at a human being’s face during an affective evaluation fundamentally alters their cognitive and psychological state, introducing a powerful, destructive psychological confound that obliterated the very somatic feedback the researchers were attempting to measure.

9.2 Objective Self-Awareness Theory as an Explanatory Framework

To provide a rigorous theoretical grounding for his critique, Strack invoked one of the most robust, deeply established theories in social psychology: Objective Self-Awareness (OSA) Theory, formulated by Shelley Duval and Robert Wicklund in 1972. Objective self-awareness theory posits that an individual’s conscious attention can be directed in one of two diametrically opposed directions:

1. Subjective Self-Awareness: Attention is focused outwardly onto the external environment and spontaneous internal experiential sensations. The individual functions as an active, experiencing subject, reacting intuitively and emotionally to external stimuli.
2. Objective Self-Awareness: Attention is actively focused backward onto the self as an object of evaluation. The individual becomes keenly aware of how they look, how they are behaving, and how they might be judged by external observers.

Duval and Wicklund, followed by decades of empirical research, proved that placing an individual in front of an external reflective or recording device—such as a mirror, an audience, or a video camera—immediately and automatically triggers an acute state of objective self-awareness. When human beings are placed under the gaze of a camera, their metacognitive monitoring systems fire into overdrive. They shift into a mode of hyper-rational, deliberate self-regulation, suppressing spontaneous bodily impulses and evaluating external stimuli through a cold, detached, socially monitored cognitive filter.

Strack argued that the facial feedback effect is an intrinsically fragile, subtle, and pre-reflective phenomenon. Proprioceptive feedback from the facial musculature operates as a low-intensity somatic whisper within the central nervous system. For this faint somatic cue to be integrated into subjective evaluations of humor, the individual must be in a state of open, relaxed, subjective experiential immersion. The moment Wagenmakers et al. mounted a video camera directly in front of participants’ faces, they forced participants into an intense state of objective self-awareness. Hyper-conscious of being filmed with an awkward plastic stick jammed in their mouths, participants’ analytical metacognitive systems completely overrode and silenced the subtle somatic feedback from the face. The camera, Strack asserted, had literally blinded the participants to their own bodily sensations.

9.3 Empirical Validation of the Camera Effect (Noah et al., 2018)

While Strack’s rebuttal was theoretically brilliant, it remained an untested hypothesis. To resolve the debate, an empirical investigation was required that directly, experimentally manipulated the presence or absence of a video camera within the exact same pen-in-mouth paradigm. This crucial empirical test was executed by Tom Noah, Yaacov Schul, and Ruth Mayo at the Hebrew University of Jerusalem, publishing their definitive findings in 2018 in the Journal of Personality and Social Psychology.

Noah, Schul, and Mayo designed a fully randomized 2 × 2 factorial experiment. They replicated the exact computerized setup, instructional scripts, and Gary Larson stimuli utilized in the 2016 Wagenmakers replication report, but systematically manipulated two independent variables:

• Facial Configuration: Teeth Condition (facilitating smile) vs. Lips Condition (inhibiting smile).
• Observation Context: Camera Present (video camera prominently mounted on the monitor, with participants told they were being recorded, exactly as in Wagenmakers et al., 2016) vs. Camera Absent (no camera, no recording, identical to Strack et al., 1988).

The results provided spectacular, undeniable empirical vindication for Fritz Strack’s theoretical defense:

• In the “Camera Present” Condition: Noah and colleagues perfectly replicated the null results of the 2016 Registered Replication Report. When participants were monitored by a video camera, there was zero statistically significant difference in cartoon funniness ratings between the teeth condition (M = 4.29) and the lips condition (M = 4.41). The effect completely vanished.
• In the “Camera Absent” Condition: When the video camera was removed and participants completed the task in private, the classic 1988 facial feedback effect returned with robust statistical power. Participants in the teeth condition rated the cartoons as significantly more amusing (M = 5.09) than participants in the lips condition (M = 4.02), yielding a statistically significant interaction (p < .05) and a large effect size (d = 0.52) that closely mirrored Strack’s original 1988 findings.

The Noah et al. (2018) study fundamentally transformed the replication narrative. The pen-in-mouth effect had not been an empirical phantom of underpowered 1980s research; rather, it was a genuine, highly sensitive psychological phenomenon that had been inadvertently eradicated in 2016 by a catastrophic, unmonitored contextual moderator: the surveillance of a video camera.

10. Modern Synthesis: The Many Faces of Facial Feedback

10.1 The Coles, Larsen, and Lench (2019) Comprehensive Meta-Analysis

The dramatic clash between the 1988 classic, the 2016 multi-lab failure, and the 2018 camera resolution highlighted the urgent need for a definitive, comprehensive quantitative synthesis of the entire scientific literature. In 2019, Nicholas Coles, Jeff Larsen, and Heather Lench published a monumental meta-analysis in Psychological Bulletin that systematically evaluated 138 empirical studies, incorporating 286 effect sizes collected from 11,211 participants across five decades of research spanning 1970 through 2019.

The Coles et al. meta-analysis was designed to answer two central questions: First, is there an authentic, statistically significant facial feedback effect across the entire body of literature, or does the effect vanish when correcting for publication bias? Second, what specific methodological moderators dictate whether the effect appears or disappears?

The meta-analytic findings provided conclusive proof that the facial feedback hypothesis is an authentic psychological reality:

• Overall Effect Size: Across all 138 studies, the meta-analysis revealed a small but statistically significant and highly reliable overall effect size of d = 0.20 (95% CI [0.14, 0.26], p < .001). While smaller than the effect originally reported by Strack in 1988, this value is fully consistent with the average effect sizes typically observed across modern social and affective psychology.
• Robustness to Bias: The researchers subjected the dataset to extensive publication-bias diagnostics, including trim-and-fill analyses, p-curve analyses, and selection models. The facial feedback effect remained statistically significant across every single sensitivity model, proving that the phenomenon is not an artifact of file-drawer effects or selective publishing.
• Moderator Verification: Crucially, the meta-analysis confirmed Noah et al.’s findings regarding social monitoring. Facial feedback effects were found to be significantly smaller, and frequently non-existent, in experimental paradigms where video cameras, mirrors, or human experimenters were directly observing the participants. Private, unobserved environments reliably yielded the strongest somatic feedback effects.

10.2 The Many Smiles Project (Coles et al., 2022)

To resolve remaining skepticism and achieve definitive consensus between opposing scientific camps, researchers initiated the Many Smiles Project, published in Nature Human Behaviour in 2022. Led by Nicholas Coles and involving a massive global consortium of over 100 researchers, the Many Smiles Project represented a landmark adversarial collaboration. It brought together ardent proponents of the facial feedback hypothesis, staunch skeptics (including key figures from the 2016 RRR), and neutral methodologists to collectively co-design and execute an absolute, definitive multi-site study.

The Many Smiles Project examined 3,878 participants across 19 countries, ensuring immense cultural and demographic diversity. The researchers decided to systematically compare three distinct, widely used facial manipulation paradigms against neutral baselines, while strictly removing all video cameras and self-focusing devices:

1. The Pen-in-Mouth Task: The classic Strack (1988) teeth vs. lips mechanical manipulation.
2. The Facial Mimicry Task: Participants viewed photographs of smiling actors and were explicitly instructed to imitate the facial expressions.
3. The Voluntary Posing Task: Participants were guided by neutral, step-by-step biomechanical instructions to contract the zygomaticus major and orbicularis oculi (e.g., “pull the corners of your lips upward toward your ears and raise your cheeks”) without ever using the word “smile.”

The results of the Many Smiles Project delivered a nuanced, fascinating theoretical resolution. The researchers discovered robust, indisputable evidence for the facial feedback effect when using both the Facial Mimicry Task and the Voluntary Posing Task. In both conditions, adopting a smiling configuration produced a statistically significant, clear increase in self-reported happiness, observed both when participants viewed neutral stimuli and when they viewed emotionally positive stimuli (p < .001). However, the classic Pen-in-Mouth Task produced ambiguous, statistically fragile results across the global sample. While the pen-in-mouth manipulation produced a tiny positive effect in some contexts, it was dramatically weaker and less reliable than deliberate facial mimicry or muscle-by-muscle voluntary posing.

10.3 Modulation versus Initiation of Affective Experience

The modern scientific consensus forged by the 2019 meta-analysis and the 2022 Many Smiles Project has profoundly clarified the precise theoretical boundaries of the facial feedback hypothesis. Crucially, contemporary affective science draws a sharp distinction between two distinct theoretical forms of feedback:

• Affect Modulation: The capacity of peripheral facial feedback to alter, amplify, or attenuate an existing, ongoing emotional state that has already been provoked by an external stimulus (e.g., making a funny cartoon appear funnier, or making an irritating film clip appear less upsetting).
• Affect Initiation: The capacity of peripheral facial feedback to spontaneously generate a completely new, discrete emotional state ex nihilo in the complete absence of any external emotional stimulus (e.g., forcing an individual sitting in an empty, neutral room into a smile and causing them to spontaneously feel genuine joy).

The cumulative evidence demonstrates that the human facial apparatus functions primarily as a modulatory amplifier rather than an absolute, autonomous initiator of emotion. While the Many Smiles Project demonstrated that voluntary facial posing can induce a slight, baseline increase in reported happiness in neutral environments, somatic feedback alone is rarely sufficient to generate an intense, fully fledged emotional experience out of thin air. Real-world human emotion is an emergent, multi-component phenomenon requiring the dynamic orchestration of subcortical arousal, cognitive appraisal, environmental context, and interoceptive feedback.

Rather than conceptualizing the face as an independent generator of emotion, modern cognitive neuroscience views facial feedback as an ongoing gain-control mechanism. When an external stimulus presents ambiguous or emotionally congruent information, the afferent somatosensory signals from the mimetic muscles provide low-level somatic evidence that tunes the perceptual and cognitive systems. If the face is smiling, the brain’s appraisal networks apply a positive interpretative gain, slightly magnifying the perceived humor or warmth of the world. If the face is scowling or pursed, the system applies an inhibitory or skeptical gain. The face does not dictate the mind’s emotional script; rather, it acts as a subtle, pervasive acoustic filter, shaping the emotional volume at which the world is heard.

11. Clinical, Neurological, and Technological Applications

11.1 Botulinum Toxin (Botox) and Affective Blunting

While experimental social psychology was fiercely debating the mechanics of the pen-in-mouth paradigm, a revolutionary, real-world neurological test of the facial feedback hypothesis was emerging from dermatological clinics: the widespread medical and cosmetic application of Botulinum Toxin Type A (Botox). Botulinum toxin acts as a potent neurotoxin that selectively cleaves SNAP-25, a critical SNARE protein required for the vesicular release of acetylcholine at the neuromuscular junction. When injected into skeletal muscle, it produces localized, temporary, and complete flaccid chemical paralysis lasting several months.

Cosmetically, Botox is most frequently injected into the corrugator supercilii and procerus muscles—the primary muscle groups responsible for drawing the eyebrows downward and together into a vertical furrow, producing the classic facial morphology of frowning, anger, sadness, and acute distress. This clinical intervention created a perfect “reverse facial feedback” paradigm: if somatic feedback from the face is causally necessary to sustain negative affective states, chemically paralyzing the frowning muscles should mechanically prevent the brain from receiving negative somatic feedback, thereby alleviating negative emotional processing.

The results of this medical line of research have provided astonishing, powerful validation for the facial feedback hypothesis:

• Neuroimaging Evidence: In a landmark fMRI study, Bernhard Haslinger and colleagues (2009) scanned female participants while they attempted to imitate angry and sad facial expressions, both before and two weeks after receiving cosmetic Botox injections into their corrugator muscles. Post-Botox scans revealed that paralyzing the frowning muscles significantly attenuated activation within the left amygdala and its functional connectivity with brainstem autonomic regions during attempted negative emotional mimicry. Deprived of peripheral corrugator feedback, the central emotional hubs of the limbic system showed profound functional blunting.
• Clinical Treatment of Major Depression: Pioneering clinical trials conducted by Tillmann Kruger, Axel Wollmer, and Eric Finzi demonstrated that a single localized injection of Botox into the corrugator muscles of patients suffering from chronic, treatment-resistant Major Depressive Disorder (MDD) produced rapid, dramatic, and sustained antidepressant effects, with clinical response rates exceeding 50% and remission rates vastly outperforming placebo saline injections. By physically abolishing the patient’s capacity to frown, the clinical intervention severed the vicious, negative somatic feedback loop maintaining depressive states in the central nervous system.

11.2 Facial Palsy, Moebius Syndrome, and Emotional Experience

Another profound clinical window into the necessity of facial feedback is provided by the study of rare neurological disorders that selectively abolish facial motor function, most notably Moebius syndrome and acquired Bell’s palsy. Moebius syndrome is an extremely rare congenital neurological condition characterized by the complete agenesis or severe underdevelopment of the sixth (abducens) and seventh (facial) cranial nerves. Individuals born with Moebius syndrome suffer from complete, lifelong bilateral facial paralysis; they cannot smile, frown, blink voluntarily, or execute any mimetic facial movements whatsoever.

Early, rigid peripheralist formulations of the facial feedback hypothesis had implied that individuals with complete facial paralysis might be entirely incapable of experiencing authentic subjective emotion. However, comprehensive neuropsychological investigations into individuals with Moebius syndrome (e.g., Bogart & Matsumoto, 2010) revealed a far more nuanced, complex reality. Individuals with Moebius syndrome experience the full qualitative spectrum of human emotions: they love, grieve, laugh internally, and experience profound personal joy. The absence of peripheral facial feedback does not abolish emotional feeling.

Nevertheless, detailed qualitative and psychometric evaluations reveal subtle, profound alterations in how emotion is processed and felt. Individuals with lifelong facial paralysis frequently describe their emotional experiences as more detached, calm, and cerebral—an internal affective state that lacks the acute, visceral, explosive intensity experienced by individuals with intact facial musculature. Furthermore, individuals who acquire sudden unilateral facial paralysis in adulthood via Bell’s palsy often report immediate, disorienting difficulties in recognizing emotions in other people. Because humans continuously utilize subtle, micro-mimetic motor simulations of observed faces to decode emotional meaning—a phenomenon known as embodied simulation—losing the motor capacity to mirror another person’s smile or frown impairs rapid, intuitive emotion recognition, forcing the individual to rely upon slower, conscious cognitive deductions.

11.3 Virtual Reality, Avatars, and Affective Computing

In the twenty-first century, the principles underlying the facial feedback hypothesis have leaped beyond human anatomy into the bleeding edge of computer science, virtual reality (VR), and affective computing. Human-computer interaction researchers are actively exploring how modulating a user’s digital or physical representation alters their internal cognitive and emotional architecture.

One of the most exciting manifestations of this research occurs via the Proteus Effect within immersive virtual environments. Modern virtual reality headsets equipped with internal infrared eye- and mouth-tracking sensors can track a user’s real-time micro-expressions and map them instantly onto their virtual avatar. Computer scientists have experimentally programmed these systems to subtly manipulate the visual feedback displayed back to the user:

• Altered Avatar Feedback: When a user looks into a virtual mirror in VR, the software algorithms can introduce a subtle visual bias, rendering the avatar’s face slightly more smiling, cheerful, and relaxed than the user’s actual physical face. Studies demonstrate that viewing one’s own avatar smiling in immersive virtual reality triggers automatic, subconscious emotional calibration: users exhibit an immediate elevation in positive mood, increased pro-social behavior, and reduced physiological stress markers.
• Haptic and Biomechanical Wearables: Researchers in affective interfaces have developed experimental wearable devices, such as customized jaw-tracking headbands and subtle electrical stimulation arrays applied to the skin above the zygomaticus major. By delivering faint, rhythmic tactile or neuromuscular micro-pulses that encourage subtle upward movement of the oral commissures during stressful tasks, these systems have succeeded in mitigating task-induced anxiety and cognitive burnout in high-stress operational environments.

These technological advancements reveal that whether mediated by an organic muscle spindle, an external wooden pen, or a digital pixel array in a virtual environment, somatic and visual feedback loops continuously sculpt human subjective reality.

12. Epistemological Legacy and the Future of Embodied Psychology

12.1 Evolution of Deception and Construct Validity in Social Psychology

From an epistemological standpoint, Fritz Strack, Leonard Martin, and Sabine Stepper’s 1988 paper remains one of the most intellectually elegant experimental paradigms ever devised in the history of behavioral science. Prior to their intervention, social and affective psychologists were routinely trapped within an empirical trade-off: they could either preserve ecological realism at the cost of crippling demand confounds, or they could enforce strict laboratory control at the cost of cognitive artificiality. Strack’s pen-in-mouth design demonstrated that a brilliant, deeply considered cover story, paired with a non-obvious biomechanical intervention, could cleanly bypass conscious cognitive appraisal, isolating an unconscious biological mechanism with extraordinary experimental elegance.

However, the historical trajectory of the pen-in-mouth paradigm also serves as a profound case study in the evolution of construct validity and research ethics. In the contemporary era of behavioral science, the use of elaborate, high-deception cover stories has faced increasing regulatory, ethical, and practical scrutiny. Institutional Review Boards (IRBs) enforce rigorous oversight regarding participant deception, while contemporary participants—raised in an era of ubiquitous digital media and psychological awareness—are notoriously more suspicious and skeptical of laboratory cover stories than undergraduate cohorts of the late 1980s. The psychomotor cover story that worked so seamlessly at the University of Mannheim in 1988 may face severe demand leakage when executed in contemporary college testing pools.

Moreover, the modern meta-analytic literature has forced psychologists to confront the delicate trade-offs inherent in physical manipulation paradigms. While Strack’s pen manipulation successfully bypassed categorical semantic labeling (the word “smile”), it introduced extraneous physical artifacts: unnatural oral sensations, saliva accumulation, and non-physiological isometric contractions. The evolution of embodied psychology has increasingly shifted away from mechanical foreign objects toward more organic, ecologically valid interventions, such as guided facial mimicry, dynamic video stimuli, and non-invasive neuroimaging, ensuring that the physical operations under investigation authentically mirror the natural physiological dynamics of the human body.

12.2 Lessons for the Metascience and Replication Movements

The epic, multi-decade saga of the pen-in-mouth hypothesis—extending from its 1988 triumph, through its 2016 multi-lab collapse, to its 2018 camera-moderator resurrection, and culminating in the 2022 Many Smiles Project—stands as one of the most instructive, iconic case studies in the history of metascience. Above all else, it delivers a profound cautionary lesson against the simplistic, binary narratives that frequently dominate scientific journalism and public discourse: the naive tendency to classify classic scientific studies as either “definitively proven” or “completely debunked.”

The 2016 Wagenmakers Registered Replication Report was a monumental triumph of open-science methodology, transparency, and statistical rigor; however, its initial public interpretation—that Fritz Strack’s 1988 finding was a historical delusion and that facial feedback does not exist—was deeply flawed. In their quest for extreme standardization and algorithmic control, the replication coordinators introduced a hidden, unmonitored methodological variable (the video camera) that fundamentally sabotaged the psychological phenomenon under study. This historical episode demonstrates that in human psychology, unlike physics or chemistry, an experimental context cannot be defined solely by computerized scripts and physical measurements. A psychological context is defined by the subjective meaning that the human participant assigns to the situation. Being secretly filmed changes everything.

The ultimate resolution of the facial feedback debate exemplifies the immense, transformative power of adversarial collaborations. Rather than retreating into entrenched, bitter academic camps—where original authors defensively reject all replication failures, and replication teams gleefully dismiss classic findings as fraudulent—the field brought proponents, skeptics, and methodologists together into a unified collaborative structure. The Many Smiles Project proved that when scientific adversaries collectively pre-register their hypotheses, debate their methodological disagreements openly prior to data collection, and share raw data transparently, the scientific process works with breathtaking self-correcting elegance.

12.3 The Lasting Scientific Contribution of Fritz Strack

Today, the enduring intellectual legacy of Fritz Strack extends far beyond the physical boundaries of a plastic pen held between human teeth. By daring to challenge the dominant, disembodied cognitive computational models of the late twentieth century, Strack helped catalyze a profound, permanent paradigm shift in how modern science conceptualizes the human mind. His work demonstrated unequivocally that human cognition, aesthetic appreciation, and subjective emotion are not abstract software programs running upon an indifferent biological computer; they are fundamentally embodied, situated, and biological processes dynamically emerging from the living flesh.

Fritz Strack’s conceptual resilience and his passionate, scientifically rigorous defense of mechanistic psychological theory ensured that the facial feedback hypothesis did not perish in the turbulent waves of the replication crisis. Instead, the paradigm emerged from the crucible of metascience infinitely more refined, precise, and neurobiologically understood. The modern synthesis acknowledges that while our facial muscles may not single-handedly dictate our subjective destinies, they are continuously whispering up the cranial nerves into the emotional control centers of our brains, gently tinting the lenses through which we perceive and evaluate our world.

Ultimately, the facial feedback hypothesis reminds us of the profound, inextricable unity of the human organism. The ancient Cartesian boundary separating the mechanical body from the thinking mind is an illusion. Every smile, whether born of spontaneous joy or summoned by the quiet, conscious mechanics of our will, ripples backward through our neurobiology, shifting the chemical currents of our brains. In the final analysis, Fritz Strack and his celebrated pen-in-mouth paradigm provided enduring empirical proof for an intuitive truth that poets and naturalists had long suspected: that the physical face we offer to the world is not merely a passive canvas upon which our internal feelings are drawn, but an active, living brush that helps paint the emotional reality of our lives.

Conclusion

The scientific odyssey of the facial feedback hypothesis—from the foundational evolutionary musings of Charles Darwin and the physiological peripheralism of William James, through Fritz Strack’s ingenious 1988 pen-in-mouth experiment, to the tumultuous trials of the replication crisis and modern adversarial collaborations—stands as a defining chapter in the history of psychological science. Strack’s landmark investigation successfully broke a historic methodological deadlock, demonstrating that the mechanical contraction of facial musculature can directly modulate complex, subjective cognitive appraisals entirely outside the realm of conscious awareness. When our facial architecture facilitates the physical morphology of a smile, the world is subtly judged as more amusing; when our facial architecture locks into an antagonistic pout, our appreciation of humor is suppressed.

While subsequent metascience initiatives—most notably the 2016 Registered Replication Report—revealed the extreme contextual fragility of this effect under the watchful gaze of modern surveillance equipment, the ultimate resolution of this academic controversy did not destroy the hypothesis. Instead, it exposed the intricate, delicate beauty of human emotional architecture. The subsequent discovery of the video camera confound, the confirmation of feedback effects in comprehensive meta-analyses, and the powerful clinical success of botulinum toxin in alleviating major depression have collectively demonstrated that the body and the mind exist in a state of continuous, bidirectional dialogue. Fritz Strack’s pen-in-mouth paradigm remains an enduring monument to scientific creativity, reminding us that within the intricate circuitry of the human nervous system, our physical actions and our emotional experiences are eternally intertwined.

References

• Bogart, K. R., & Matsumoto, D. (2010). Facial mimicry is not necessary to recognize emotion: Facial expression recognition by people with Moebius syndrome. Social Psychological and Personality Science, 1(3), 241–247. https://doi.org/10.1177/1948550610363484
• Cannon, W. B. (1927). The James-Lange theory of emotions: A critical examination and an alternative theory. The American Journal of Psychology, 39(1/4), 106–124. https://doi.org/10.2307/1415404
• Coles, N. A., Larsen, J. T., & Lench, H. C. (2019). A meta-analysis of the facial feedback literature: Effects of facial feedback on emotional experience are small and variable. Psychological Bulletin, 145(6), 610–651. https://doi.org/10.1037/bul0000194
• Coles, N. A., March, D. S., Marmolejo-Ramos, F., Larsen, J. T., Arnal, J. D., Ndukaihe, I. L. G., … & Liuzza, M. T. (2022). A multi-lab test of the facial feedback hypothesis by the Many Smiles Collaboration. Nature Human Behaviour, 6(12), 1731–1742. https://doi.org/10.1038/s41562-022-01458-9
• Darwin, C. (1872). The expression of the emotions in man and animals. John Murray. https://doi.org/10.1037/10001-000
• Duval, S., & Wicklund, R. A. (1972). A theory of objective self awareness. Academic Press.
• Ekman, P. (1992). An argument for basic emotions. Cognition & Emotion, 6(3–4), 169–200. https://doi.org/10.1080/02699939208411068
• Finzi, E., & Rosenthal, N. E. (2014). Treatment of depression with onabotulinumtoxinA: A randomized, double-blind, placebo controlled trial. Journal of Psychiatric Research, 52, 1–6. https://doi.org/10.1016/j.jpsychires.2013.11.006
• Haslinger, B., Aldinger, F., Zimmer, C., & Castrop, F. (2009). Modulation of amygdala response by facial feedback: An fMRI study using botulinum toxin. European Journal of Neuroscience, 29(5), 1045–1052. https://doi.org/10.1111/j.1460-9568.2009.06648.x
• Izard, C. E. (1971). The face of emotion. Appleton-Century-Crofts.
• James, W. (1884). What is an emotion? Mind, 9(34), 188–205. https://doi.org/10.1093/mind/os-IX.34.188
• Laird, J. D. (1974). Self-attribution of emotion: The effects of expressive behavior on the quality of emotional experience. Journal of Personality and Social Psychology, 29(4), 475–486. https://doi.org/10.1037/h0036125
• Noah, T., Schul, Y., & Mayo, R. (2018). When both the original study and its replication are correct: The modern video-camera transform the pen-in-mouth effect. Journal of Personality and Social Psychology, 114(5), 657–664. https://doi.org/10.1037/pspa0000121
• Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45(3), 513–523. https://doi.org/10.1037/0022-3514.45.3.513
• Strack, F. (2016). Reflection on the smiling registered replication report. Perspectives on Psychological Science, 11(6), 929–930. https://doi.org/10.1177/1745691616674460
• Strack, F., Martin, L. L., & Stepper, S. (1988). Inhibiting and facilitating conditions of the human smile: A nonobtrusive test of the facial feedback hypothesis. Journal of Personality and Social Psychology, 54(5), 768–777. https://doi.org/10.1037/0022-3514.54.5.768
• Tomkins, S. S. (1962). Affect imagery consciousness: Vol. I. The positive affects. Springer Publishing Company.
• Tomkins, S. S. (1963). Affect imagery consciousness: Vol. II. The negative affects. Springer Publishing Company.
• Tourangeau, R., & Ellsworth, P. C. (1979). The role of facial response in the experience of emotion. Journal of Personality and Social Psychology, 37(9), 1519–1531. https://doi.org/10.1037/0022-3514.37.9.1519
• Wagenmakers, E. J., Beek, T., Dijkhoff, L., Gronau, Q. F., Acosta, A., Adams, R. B., … & Zwaan, R. A. (2016). Registered replication report: Strack, Martin, & Stepper (1988). Perspectives on Psychological Science, 11(6), 917–928. https://doi.org/10.1177/1745691616674458
• Wollmer, M. A., de Boer, C., Kalak, N., Beck, J., Götz, T., Schmidt, T., … & Kruger, T. H. (2012). Facing depression with botulinum toxin: A randomized controlled trial. Journal of Psychiatric Research, 46(5), 574–581. https://doi.org/10.1016/j.jpsychires.2012.01.027
• Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35(2), 151–175. https://doi.org/10.1037/0003-066X.35.2.151
• Zajonc, R. B., Murphy, S. T., & Inglehart, M. (1989). Feeling and facial efference: Implications of the vascular theory of emotion. Psychological Review, 96(3), 395–416. https://doi.org/10.1037/0033-295X.96.3.395