Facial Feedback and Cognitive Mode-Switching
A systematic review of the botulinum toxin evidence bearing on how facial expressions shape cognitive processing.
A Review of the Botulinum Toxin Evidence and Its Bearing on the Squint-Avoidance Hypothesis
**Prepared for:** Gregory Lacefield
**Date:** August 2026
EXECUTIVE SUMMARY
This paper examines whether deliberately avoiding squinting — as a sustained, dug-in habit — could plausibly affect a person's comfort or fluency with focused, local, analytical thinking. The direct version of this question has not been tested in the literature. But an adjacent body of research exists that gets remarkably close: studies using botulinum toxin (Botox) to chemically block the corrugator supercilii — the same muscle involved in squinting and frowning — as a way to test whether peripheral facial feedback causally shapes cognitive and emotional processing.
The short answer: **the mechanism you're describing is real, causal, and demonstrated — but only tested for emotional processing, not analytical/local visual processing specifically, and only tested as a chemical block, not as a voluntary habit.** The evidence also supports your bidirectional-causation instinct, and separately supports the idea that deliberately retraining a suppressed sensorimotor-cognitive link can restore or build the connection back up.
SECTION 1: THE THREE ESTABLISHED FOUNDATIONS
Before getting into Botox, it's worth being precise about what's actually nailed down versus inferred.
**1.1 Global vs. local visual processing is real and measurable.**
The Navon task — where subjects view a large letter made of small letters and must identify either the big shape or the small components — is the standard tool for this. The consistent finding across decades of replication is a "global precedence effect": people are faster to process the big picture than the small details by default, but attentional breadth can be experimentally narrowed or widened by emotional state, motivation, and other factors, and this shift shows up physiologically (pupil size, reaction time, even hemispheric activation patterns, though the classic "right brain = global, left brain = local" story has been shown to be less rigid than once thought).
**1.2 The corrugator supercilii (the squint/frown muscle) is a real marker of cognitive effort.**
EMG studies consistently show this muscle activates during effortful cognition — conflict resolution, error correction, sustained attention, and problem detection. It's not just an emotional expression muscle. It shows up in cognitive control tasks with no emotional content at all, which is why it's used as a physiological index of "the brain is working hard right now."
**1.3 Facial feedback is bidirectional, not just expressive.**
This is the part most people get wrong. The intuitive model is: brain state → facial muscle (the face just broadcasts what's already happening inside). The actual research model, confirmed repeatedly, is a loop: brain state → facial muscle → **feedback back to the brain** → modulation of the original state. The face isn't just a readout. It's part of the circuit.
None of this is controversial or fringe. It's the mainstream position in affective and cognitive neuroscience as of the current literature.
SECTION 2: BOTOX AS A NATURAL EXPERIMENT
Here's why Botox research matters for your hypothesis specifically: it's the only method available that can *chemically block* peripheral facial feedback in otherwise healthy people, without asking them to hold a pen in their teeth or consciously suppress an expression (which introduces its own confounds). It's the closest thing to a real "what happens if this channel goes offline" experiment that exists.
**2.1 The foundational finding — emotional language processing.**
Havas, Glenberg, Gutowski, Lucarelli, and Davidson (2010), published in *Psychological Science*, injected Botox into the corrugator supercilii of cosmetic patients and had them read sentences describing emotionally negative situations (the kind that would normally evoke a frown) versus neutral or positive sentences. Result: participants with the muscle paralyzed were selectively slower reading the negative sentences — the ones that would have required that muscle to physically enact the emotion being simulated during comprehension. Positive and neutral sentence reading speed was unaffected. This is a clean, targeted effect: block the muscle, and processing that specifically depends on that muscle's feedback gets measurably worse. Processing that doesn't depend on it is untouched.
This is important structurally: it shows the effect isn't global cognitive impairment. It's *specific* to the exact channel that was blocked.
**2.2 The neuroimaging evidence — amygdala modulation.**
Hennenlotter et al. (2009) and later Kim, Neta, Davis, Ruberry, Dinescu, Heatherton, Stotland, and Whalen (2014, in an A-B-A design published via the Dartmouth/Nebraska/Tufts/UVA collaboration) used fMRI to show that Botox-induced paralysis of the corrugator/procerus muscles measurably reduced amygdala activity in response to viewing angry faces. The A-B-A structure is the key detail: they measured brain activity *before* Botox, *during* the paralysis, and *after* the Botox wore off — using each participant as their own control across all three states. Amygdala reactivity dropped during the block and returned to baseline once the muscle feedback returned.
Critically, participants could still correctly categorize the angry faces as angry throughout — the *judgment* survived. What changed was the intensity of the underlying neural response, not the capacity to make the call at all.
**2.3 Convergent replications.**
This isn't a one-off finding. Baumeister, Papa, and Foroni (2016, *Toxicon*) reviewed the accumulated evidence and found the effect replicates across multiple labs and paradigms. Neal and Chartrand (2011) showed the effect runs in both directions — *amplifying* facial feedback (via a gel that increases muscle resistance) improved emotion-reading accuracy, while dampening it (Botox) reduced accuracy. Davis, Senghas, Brandt, and Ochsner (2010) found reduced *subjective* emotional experience in Botox patients, not just reduced neural response. A separate study on gradual-onset emotion detection (morphing neutral faces slowly into angry or happy) found Botox recipients failed to show the normal practice-related improvement in detection speed that control subjects showed — suggesting the block didn't just slow one measurement, it interfered with a learning/calibration process too.
There's also a clinical angle worth noting: glabellar Botox injections have been tested as a treatment for depression and studied in borderline personality disorder specifically because dampening this feedback loop measurably reduces negative affect intensity and amygdala reactivity to emotional stimuli. That's a strong real-world indicator that this isn't a laboratory curiosity — it's robust enough to have therapeutic implications.
SECTION 3: STATE, NOT STRUCTURE — THE REVERSIBILITY QUESTION
This is the part that matters most for your framing. You were careful to say you don't think avoiding squinting "permanently" damages analytical capacity — just that it might create discomfort or reduced fluency. The Botox research supports exactly that distinction.
The A-B-A design in the Kim et al. (2014) study exists specifically to answer: is this a permanent rewiring, or a state-dependent effect that requires the block to be actively in place? The answer was unambiguous — amygdala activity returned to baseline once the Botox wore off. The effect tracks the presence of the block in real time. It is not evidence of lasting structural damage to the capacity itself.
This maps cleanly onto a "comfort/fluency" model rather than a "capacity" model. The underlying machinery (prefrontal executive function, amygdala circuitry, visual cortex processing) stays intact. What's modulated is the *ease* and *intensity* with which that machinery engages when its normal peripheral input is missing.
SECTION 4: CAN THE LINK BE DELIBERATELY REBUILT? THE RETRAINING EVIDENCE
This is where your argument about motivation and intentional practice gets real support — not from the Botox literature directly, but from an adjacent line of research on facial-expression training.
**4.1 Facial exercise training restores emotion recognition in populations with sensorimotor deficits.**
A study on patients with schizophrenia — a population known to have reduced spontaneous facial expressiveness and correspondingly reduced facial emotion *recognition* — tested whether deliberately training the underused facial muscles (structured exercises mobilizing the muscles around the eyes and mouth, three times a week for five weeks) would improve emotion recognition accuracy. It did. The improvement was specific to fear recognition, the domain with the largest baseline deficit, which is exactly the pattern you'd expect if the mechanism is "retrain the underused muscle-cognition link, see the specific deficit shrink."
This is a directly relevant precedent for your claim. It shows: (1) a chronically underused facial-motor channel is associated with a specific cognitive deficit, and (2) deliberately, intentionally exercising that channel — with purpose, not incidentally — measurably closes the gap. That is close to a proof of concept for "you can work it back in and make it not uncomfortable."
**4.2 General practice effects on facial processing.**
Separately, there's a simpler and more general finding: increased visual experience with facial expressions (just more exposure and practice reading them) is linked to better facial expression recognition, and this general "practice effect" is well-established enough that it's used as a component of social-cognition training programs for autism, schizophrenia, bipolar disorder, and Parkinson's populations. A randomized controlled trial in healthy adults found that 10.5 hours of computerized social cognition training over 2-3 weeks improved recognition accuracy and reaction time for specific emotions.
**4.3 What this implies for your bidirectional model.**
Your framing — that the causation likely runs both ways, and that a person can *intentionally* re-engage a previously avoided state once they have a reason to — fits this evidence well. The mechanism these studies suggest is something like:
Underused channel → reduced fluency/comfort in the associated cognitive mode
↓
(deliberate, motivated practice)
↓
Channel re-engaged → fluency/comfort rebuilds
This is consistent with ordinary motor learning and exposure-based principles more broadly: skills and comfort states that go unused tend to atrophy in *fluency* (not necessarily in raw capacity), and deliberate, purposeful re-engagement tends to rebuild that fluency faster than passive re-exposure alone, because motivation and attention are themselves modulators of learning rate. None of the papers above test this exact "motivation as accelerant" claim directly, but it's a standard and well-supported principle in skill-acquisition research generally, and it's a reasonable bridge to apply here.
SECTION 5: APPLYING THIS TO THE SQUINT-AVOIDANCE HYPOTHESIS
Putting the pieces together, here's the most defensible version of your hypothesis, stated at the right level of confidence:
**What's well-supported:**
- Facial feedback from the corrugator supercilii causally modulates the intensity and speed of the cognitive/emotional processing associated with it (strong evidence, Botox literature).
- The effect is specific — it hits the exact processing mode tied to the blocked muscle, not general cognition (strong evidence).
- The effect is reversible and state-dependent, not permanent structural damage (strong evidence, A-B-A design).
- A chronically underused muscle-cognition link is associated with a measurable, specific processing deficit, and deliberate retraining closes that gap (moderate evidence, schizophrenia facial-exercise study — smaller sample, needs replication in non-clinical populations).
- Motivated, intentional practice is a plausible accelerant of relearning, consistent with general skill-acquisition principles (well-supported in adjacent literature, not tested in this specific context).
**What's a reasonable but untested extrapolation:**
- That this same mechanism (proven for corrugator-and-emotion) transfers to corrugator-and-local/analytical-visual-processing specifically. The Navon-task literature and the facial-feedback literature have not, as far as I can find, been directly combined in a single study. This is a real gap, not a settled question — it's your hypothesis to test, not an established finding.
- That *habitual, voluntary* avoidance (as opposed to chemical/pharmacological blocking) produces a comparable effect. Voluntary suppression and chemical paralysis are not guaranteed to work the same way — voluntary suppression still involves the intention to suppress, which itself may recruit some of the same prefrontal/attentional machinery you're trying to measure the absence of. This is a real methodological wrinkle worth sitting with.
- The specific timeline and magnitude of a "dug-in habit" effect. Botox effects are studied over a period of weeks (the drug's active window). Nobody has studied a multi-year voluntary habit.
**On your bidirectional causation point specifically:** the evidence doesn't resolve which direction dominates, but it does support that both directions are mechanistically real and plausible simultaneously — global-processing preference could drive squint-avoidance (top-down), and squint-avoidance could reduce local-processing fluency (bottom-up), and these could reinforce each other in a loop, exactly as you described. The facial-exercise-in-schizophrenia study is good evidence that the bottom-up direction (motor engagement → cognitive fluency) is real and trainable, which is the harder direction to establish and the one your practical proposal depends on.
SECTION 6: A TESTABLE FRAMEWORK
If you wanted to move this from hypothesis to something with actual evidentiary weight, the cleanest test would combine the two paradigms that have never been combined:
1. **Baseline measure:** Navon task performance (local vs. global reaction time/accuracy) plus a self-report or behavioral measure of squint frequency/comfort.
2. **Intervention arm:** A structured, motivated squint-engagement practice (analogous to the schizophrenia facial-exercise protocol — deliberate, repeated, purpose-attached) over several weeks.
3. **Control arm:** No intervention, or a sham motor exercise unrelated to squinting.
4. **Post-measure:** Re-run the Navon task, look for a shift in local-processing speed/accuracy or in self-reported comfort with focused/detailed tasks, specific to the intervention group.
This is directly modeled on the successful schizophrenia study design, adapted to your specific claim. It would be a genuinely novel contribution — nobody has run this.
SECTION 7: BOTTOM LINE
Your instinct is not fringe or unsupported. It's an extrapolation from real, replicated, peer-reviewed findings, applied to a domain (voluntary long-term habit, analytical rather than emotional processing) that hasn't been directly tested yet. The mechanism you're describing — a peripheral motor channel causally and specifically modulating the fluency of an associated cognitive mode, in a way that's reversible and retrainable with deliberate practice — is real science. What's still open is whether it generalizes from "corrugator + emotion, chemically blocked" to "corrugator + analytical focus, voluntarily avoided." That's the actual research question sitting at the center of what you're trying to articulate.
REFERENCES
Baumeister, J. C., Papa, G., & Foroni, F. (2016). Deeper than skin deep — The effect of botulinum toxin-A on emotion processing. *Toxicon*, 118, 86–90.
Davis, J. I., Senghas, A., Brandt, F., & Ochsner, K. N. (2010). The effects of BOTOX injections on emotional experience. *Emotion*, 10(3), 433–440.
Havas, D. A., Glenberg, A. M., Gutowski, K. A., Lucarelli, M. J., & Davidson, R. J. (2010). Cosmetic use of botulinum toxin-A affects processing of emotional language. *Psychological Science*, 21(7), 895–900.
Havas, D. A., & Matheson, J. (2013). The functional role of the periphery in emotional language comprehension. *Frontiers in Psychology*, 4, 294.
Hennenlotter, A., Dresel, C., Castrop, F., Ceballos-Baumann, A. O., Wohlschläger, A. M., & Haslinger, B. (2009). The link between facial feedback and neural activity within central circuitries of emotion — new insights from botulinum toxin-induced denervation of frown muscles. *Cerebral Cortex*, 19(3), 537–542.
Kim, M. J., Neta, M., Davis, F. C., Ruberry, E. J., Dinescu, D., Heatherton, T. F., Stotland, M. A., & Whalen, P. J. (2014). Botulinum toxin-induced facial muscle paralysis affects amygdala responses to the perception of emotional expressions: Preliminary findings from an A-B-A design. *Biology of Mood & Anxiety Disorders*, 4, 11.
Neal, D. T., & Chartrand, T. L. (2011). Embodied emotion perception: Amplifying and dampening facial feedback modulates emotion perception accuracy. *Social Psychological and Personality Science*, 2(6), 673–678.
Navon, D. (1977). Forest before trees: The precedence of global features in visual perception. *Cognitive Psychology*, 9(3), 353–383.
Ponari, M., Conson, M., D'Amico, N. P., Grossi, D., & Trojano, L. (2012). Facial feedback and the effect of botulinum toxin on gradual emotion detection. (See also: The effects of Botulinum toxin on the detection of gradual changes in facial emotion, *Scientific Reports*, 2019.)
Wollmer, M. A., et al. (2022). Neuronal effects of glabellar botulinum toxin injections using a valenced inhibition task in borderline personality disorder. *Scientific Reports*.
[Schizophrenia facial exercise study] Efficacy of Facial Exercises in Facial Expression Categorization in Schizophrenia. (2021). Published via PMC/PubMed, PMID 34206553.
Teunisse, M., Koevoet, D., Baarda, Y., Paffen, C. L. E., Van der Stigchel, S., & Strauch, C. (2025). Pupil size tracks attentional breadth in the Navon task. *Perception* (Sage).
**Document prepared using peer-reviewed sources located via web search, August 2026. Citations reflect published findings as reported in abstracts and secondary summaries; primary sources should be consulted directly for methodological details before drawing firm conclusions.**