The voluntary control of piloerection

Introduction

“Certainly that heart is steel-framed which, in spite of one’s chanting the holy name of the Lord with concentration, does not change when ecstasy takes place, tears fill the eyes and the hairs stand on end.”

Bhãgavata Purãna, 2.3.24, 800-1000AD  (Bhaktivedanta VedaBase, 2018).

The study of exceptional individuals is central to the history of human neuroscience. Famous case studies include patients with specific neurological insults, such as naturally occurring lesions (e.g., Louis Leborgne; Broca, 1861), victims of crimes or accidents (e.g., Phineas Gage; Harlow, 1848), or patients receiving neurosurgery (e.g., Henry Melaison; Corkin, 1984). Detailed observations of these individuals were critical to initial developments within language processing, behavioural inhibition, and amnesia/memory consolidation respectively. Non-invasive imaging techniques allow contemporary research to pursue similar case study or small-sample approaches without autopsy, and relevant patients continue to make important contributions to the neuroscience of behaviour. For instance, patients B.G. and A.M., who have bilateral amygdalae lesions, show a differential panic response after a panic induction procedure, demonstrating that the amyagdala may not be necessary for the induction of panic (Khalsa et al., 2016). Inspired by this tradition, the present investigation seeks to identify, describe, and compare a small sample of people with atypical autonomic physiology who are otherwise healthy to identify corresponding patterns in their physical, cognitive, and emotional experiences.

An example here is illustrative: suppose two people, one healthy and neurotypical, the other identical but with a specific injury to the thoracic sympathetic ganglia. If the first walks face-first into a spider’s web in the dark, we might expect subsequent reflex reactions (e.g., eye-blink, self-defensive posture) followed by the autonomic nervous system (ANS) increasing heart rate (HR) in preparation for whole-body mobilisation. The second would most likely have typical reflexes, including rapid appraisal of the situation (i.e., a small but definite cutaneous sensation on the face), and then subsequently a normal cognitive response (i.e., recognition of the presence of a spider, generally categorized by humans as dangerous and/or disgusting), but reduced physical symptoms of fear or panic, as their heart remains steady, palms dry, and so on. That is, this ‘sympathetically-impaired’ person would experience a dissociation between his or her appraisal of what is typically fear-inducing and the typical embodied experience of fear. If we consider emotion an interface between cognitive interpretations and physical experience of the body (e.g., Schachter & Singer, 1962; Damasio, 1999; Barrett, 2017), how should we understand perception of the relevant physiology being altered? What then would the experience be?

As an additional layer of complication, consider the above scenario in an individual with strong conscious control over their HR. How does one experience an embodied mental state when he or she can consciously control its autonomic precedents or antecedents? Somewhat paradoxically, the relevant control in this case would be over the ANS, the portion of the peripheral nervous system traditionally defined as operating outside volitional or conscious control, from the Greek autos (self) and nomos (law), i.e., self-governing. However, delineating autonomic and somatic (i.e., voluntary) functions is not clear-cut. Respiration is the classic example of an autonomic mechanism directly amenable to conscious control. Absent any form of conditioning, there is also evidence of volitional control of HR (Bell & Schwartz, 1975; White, Holmes & Bennett, 1977) and blood pressure (Lowdon, Murray & Langley, 2011). As autonomic systems co-modify, the above effects may be combined. For instance, conscious control of respiration might also be considered a deliberate modification of blood pressure (Joseph et al., 2005). Some curious case studies from meditative traditions concerning HR control (Young & Taylor, 1998) and temperature (Benson et al., 1982) have also been reported. Related research has additionally taken place under the rubric of biofeedback, where conscious access to a previously non-conscious ability is strengthened by physiological instrumentation (e.g., somatic control of skin temperature; Taub & School, 1978). Finally, a separate case again is the conscious control of pupillary contraction and dilation, which is both capable of being conditioned (Cason, 1922) and responds directionally to imagined changes in luminance (Laeng & Sulutvedt, 2014).

In addition to the above, there is an autonomic mechanism yet to be systematically investigated in terms of volitional control –piloerection, often referred to as ‘goosebumps.’ Involuntary piloerection is commonly observed during cold, fearsome, or intensely evocative emotional stimuli, and is subserved by the pilomotor projections of the sympathetic nervous system (SNS). However, while evidence in the scientific literature suggesting piloerection can be consciously controlled has been observed in only three published case studies over a period of more than a century (Benedek et al., 2010; Lindsley & Sassaman, 1938; Maxwell, 1902), the incidence of this ability, method of action, and potential psychological correlates have not been thoroughly examined. Voluntarily generated piloerection (VGP) is described below in brief (Table 1).

In light of other examples of conscious physiological control, VGP is unusual for several reasons. First, the arrector pili which control erection of individual hairs are smooth muscles—they have no somatic control in the manner of skeletal muscles, such as the biceps or quadriceps. Rather, they operate outside of conscious awareness, contracting or relaxing to local cellular factors or the general neurochemical environment. A reductionist physiological account would conclude conscious piloerection is not possible. Second, VGP was described in all three prior case studies as being ‘discovered’ by the person experiencing it. That is, they all outline various experiences where VGP was uncovered as a latent ability without conditioning, biofeedback, expectation, or training. Thus, no pathway to acquire this ability is known. Third, VGP offers control over a physiological phenomenon intimately involved in emotional experience, wherein bodily sensations, emotional states, and experiential terminology frequently overlap (‘her palms were damp,’ ‘he had butterflies in his stomach,’ and ‘their hearts were pounding’), and likewise here ‘my hair was standing on end.’

Both classic and modern accounts of the nature of emotion consider the integrated role of perception and feedback of somatic symptoms such as goosebumps (e.g., Schachter & Singer, 1962; Damasio, 1999; Barrett, 2017). Other self-perceived physical symptoms, such as the ability to feel the heart beating in the chest (‘cardiac interoception’), associate with emotional (Pollatos, Kirsch & Schandry, 2005) and physical (Herbert, Ulbrich & Schandry, 2007) self-regulation. Involuntary piloerection has primarily been studied in the context of intense emotional experiences, has been shown to correlate strongly with reported emotional intensity, and seems to be particularly associated with states of awe (Schurtz et al., 2012) or being ‘moved’ or ‘touched’ (Benedek & Kaernbach, 2011; Wassiliwizky et al., 2015). Whether VGP relates in any way to subjective experiences associated with involuntary piloerection remains an open question. Likewise, relationships between VGP and the tendency to experience involuntary piloerection is unknown. Finally, very little is known about psychological correlates of VGP. Is it accompanied by emotions and sensations usually associated with involuntary piloerection? Do people capable of VGP use this ability to enhance or moderate their psychological experiences?

Only one of the three case studies cited above mentions any psychological aspect; the participant in question reported that voluntary piloerection was experienced as “rather pleasurable than otherwise” (Maxwell, 1902, p. 373). This individual further reported unusual features—he could use VGP for headache relief, observed no special sensitivity during fear, sympathy, or music, and experienced involuntary piloerection most of the time while watching football (during which VGP was reportedly unusually easy). The other case studies primarily focus on automated detection or physiology of VGP (Benedek et al., 2010; Lindsley & Sassaman, 1938).

Another question of interest is whether people capable of VGP have unique personality profiles. Maxwell (1902) initially speculated that the ability may associate with a neurotic personality disposition. Since then, individual differences have been found in the propensity to experience involuntary piloerection and related psychological states and are most consistently associated with the personality domain of openness to experience—a domain reflecting variability in cognitive exploration (DeYoung, 2015). A questionnaire item reporting on experiences of aesthetic chills in response to music and poetry is one of the highest loading items on the openness to experience questionnaire, suggesting that experiences of aesthetic chills are a universal marker for openness to experience (McCrae, 2007). Furthermore, openness associates with chills in response to music (Colver & El-Alayli, 2016; Nusbaum & Silvia, 2011; Silvia et al., 2015), and the arts more generally (Silvia & Nusbaum, 2011). Open people are also found to be more prone to being absorbed in their experiences generally (Wild, Kuiken & Schopflocher, 1995), and in aesthetic contexts specifically (Silvia & Nusbaum, 2011). They are also more likely to report experiences of awe (Shiota, Keltner & John, 2006), and recall awe-based crying experiences with music (Cotter, Silvia & Fayn, 2018). Given the lack of consistent relationships between involuntary piloerection or chills and other personality domains, no hypotheses were formulated for other personality domains. However, given that individual differences in VGP have never been previously explored, a short measure of the Big Five was administered for exploratory comparisons with normative data.

Much of this prior work addresses the relationship between openness and experience of ‘chills.’ While chills are not identical to involuntary piloerection in the sense that they may not necessary involve either body hair moving visibly or the skin dimpling due to activity of the arrector pili, and they have distinct physiological correlates (Sumpf, Jentschke & Koelsch, 2015; Wassiliwizky et al., 2017b) experiences of chills and involuntary piloerection occur in response to similar situations (music, poetry, art) and in concordance with similar emotional experiences (awe, being moved). Therefore, we hypothesize that openness associates with a greater propensity to experience both chills and involuntary piloerection, though to our knowledge, no study has previously explored this possibility.

The links between involuntary piloerection, the experience of chills, and existing personality correlates suggest that individuals with voluntary control over these symptoms tend to experience states of being moved, touched, and/or awed more frequently, and score higher on openness to experience measures. However, hypothesised differences in openness are contingent on whether the experience of VGP is accompanied by psychological experiences associated with involuntary piloerection. In summary, VGP provides potential insight into both the nature of autonomic regulation and a correlate of emotion and personality. The present study is the first we are aware of to characterise this ability in a sample consisting of more than one individual.

Method

Results

Psychological features of VGP

Most participants (71.88%) reported at least one psychological state previously found to relate to involuntary piloerection accompanying their VGP experience. The most common reported were absorption/immersion (53.13%), awe/wonder (46.88%), and detachment (37.5%). The least commonly reported were feeling: ‘touched,’ ‘being somewhere else,’ ‘crying,’ or ‘losing track of time.’ The three most endorsed items are all part of the absorption/immersion subscale of the aesthetic experiences scale reflecting the propensity to experience powerful states of engagement and awe (Silvia & Nusbaum, 2011), suggesting that feelings accompanying VGP are similar to reports of awe-induced involuntary piloerection (Schurtz et al., 2012). Since much of the recent work on involuntary piloerection links the phenomenon to a state of being moved (Wassiliwizky et al., 2017a; Wassiliwizky et al., 2017b), one might expect our sample to also endorse items in the touched subscale of the aesthetic experience scale, but this was not the case. From one perspective, this may indicate that experiences during VGP are distinct from experiences of being moved. On the other hand, it is difficult to imagine feelings of being moved in the absence of a stimulus that moves you, as is imagining being moved without being absorbed by the stimulus to some degree. Participants also reported the following normal cutaneous sensation of goosebumps during commonly piloerective stimuli: when cold, during music, during films or television, when afraid, when touched lightly, during the experience of aesthetic or natural beauty, etc. (note: these are in decreasing rank order).

Participants generally reported to have no personal explanation for their ability (37.5%) or offered a scientific mechanism (37.5%; descending noxious inhibitory control, various neurotransmitters, the autonomic nervous system, circulating epinephrine, biofeedback, etc.) Only a single participant mentioned anything akin to a spiritual or non-mechanistic process.

Next, we asked whether participants used their VGP during activities that tend to elicit involuntary piloerection. Of the sample, 71.88% reported using their ability in response to at least one of the activities. The most frequently reported activities where participants used their ability were music (53.13%), film/television (28.13%), literature (25%), creative activities (25%), art (21.88%), dance (18.75%), theatre (15.63%), and gaming (12.5%). The ‘other’ category was also frequently endorsed (46.88%), and included meditation, engagement with nature, exercise, sex, work or study for added energy and concentration, and to warm up. Only one participant mentioned using the ability to alleviate pain, as was the case in Maxwell (1902). Thus, most of the sample reported using their ability in contexts where involuntary piloerection is frequently reported. Around half (53.13%) reported deliberately using their ability to prolong experiences of involuntary piloerection.

To test whether VPG-capable individuals experience emotions and experiences associated with involuntary piloerection more frequently, we compared their scores on the aesthetic experiences scale to the largest available sample previously collected from university students (n = 188; Silvia & Nusbaum, 2011). VGP-capable participants (M = 4.75, SD = 1.50), compared to a previously collected sample (M = 4.55³ ), reported no difference in the frequency of feeling absorbed (mean difference = .20; t(27) = 0.69, p = .50, Cohen’s d = .13, 95% CI [−.39–.78]), and no difference (M = 3.98, SD = 1.67 versus M = 3.63) in the frequency of chills (mean difference = .26; t(29) = 1.14, p = .26, Cohen’s d = .21, 95% CI [−.27, .97]), and no difference (M = 3.55, SD = 1.56 versus M = 4.13) in being moved (mean difference = −.58; t(29) =  − 2.02, p = .053, Cohen’s d = .37, 95% CI [−1.16–.08]).

Since we had no prior information regarding sample characteristics of our VGP-capable participants, we compared personality data to normative data from the largest published samples for the relevant scales (Table 2). The VGP sample was significantly higher on BFAS Openness/Intellect, in both the domain and aspect level, and the differences produced large effect sizes. This was not replicated with the TIPI measure of openness, but the difference was significant for one of the openness items. No other differences were observed at the domain level of the TIPI, but there were some differences at the item level. However, none met Bonferroni-corrected threshold for significance (α∕10 = 0.005). The VGP sample scored significantly lower on an agreeableness item, and significantly higher on an emotional stability item; both differences were medium in effect size.

Table 2:

VGP sample vs. population norms for Ten-Item Personality Inventory items and domains, Big Five Aspect Scale subscales and domain.

Personality measure	Sample		Norms		Difference		Confidence intervals
TIPI Items	Mean	SD	Mean	SD	p	Cohen’s d
Extraverted, enthusiastic.	4.73	1.80	4.42	1.73	0.354	0.17	[−0.19, 0.53]
Critical, quarrelsome (R).	3.43	1.72	4.08	1.72	0.047∗	−0.38	[−0.75, 0.01
Dependable, self-disciplined.	5.50	1.36	5.00	1.54	0.055	0.37	[−0.01, 0.73]
Anxious, easily upset (R).	4.20	1.81	4.06	1.81	0.671	0.08	[−0.28, 0.44]
Open to new experiences, complex.	6.00	1.20	5.49	1.34	0.029∗	0.42	[0.04, 0.79]
Reserved, quiet (R).	4.10	1.58	3.54	1.88	0.063	0.35	[−0.02, 0.72]
Sympathetic, warm.	5.43	1.14	5.29	1.44	0.502	0.12	[−0.24, 0.48]
Disorganized, careless (R).	4.40	1.65	4.22	1.85	0.552	0.11	[−0.25, 0.47]
Calm, emotionally stable.	5.20	1.54	4.62	1.64	0.049∗	0.38	[0, 0.74]
Conventional, uncreative (R).	5.23	1.78	5.52	1.54	0.385	−0.16	[−0.52, 0.20]
TIPI Domain
Extraversion	4.42	1.46	3.98	1.59	0.114	0.30	[−0.07, 0.66]
Agreeableness	4.43	1.21	4.69	1.23	0.257	−0.21	[−0.57, 0.15]
Conscientiousness	4.95	1.29	4.61	1.42	0.161	0.26	[−0.10, 0.62]
Emotional Stability	4.70	1.32	4.34	1.48	0.145	0.27	[−0.09, 0.64]
Openness	5.62	1.15	5.51	1.14	0.604	0.10	[−0.26, 0.45]
BFAS
Openness/Intellect	4.05	0.48	3.60	0.51	0.00002∗	0.93	[0.50, 1.36]
Openness	4.03	0.50	3.68	0.61	0.0005∗	0.72	[0.31, 1.11]
Intellect	4.07	0.68	3.53	0.63	0.0001∗	0.8	[0.39, 1.21]

DOI: 10.7717/peerj.5292/table-2

Notes:

Reversed items are marked (R) and presented recoded in line with domain label.

As a post-hoc follow-up, Fig. 1 compares the BFAS Openness/Intellect omnibus score against large samples where relevant data was available (8 samples drawn from six papers; Antinori, Carter & Smillie, 2017; DeYoung, Quilty & Peterson, 2007; Fayn et al., 2017; Fayn et al., 2015; Sun, Kaufman & Smillie, 2018; Weisberg, DeYoung & Hirsh, 2011). Uncorrected p-values (t-test, two-tailed) range from 0.051 to 10⁻⁷, and effect sizes range from Cohen’s d = 0.35 to 1.16.

Discussion

To our knowledge, this paper outlines the first sizable sample characterised who report having a voluntary ability to produce goosebumps. While this sample was not observed directly in a laboratory setting, participants principally reported an experience consistent with those previously identified in laboratory case-studies over the previous century (Benedek et al., 2010; Lindsley & Sassaman, 1938; Maxwell, 1902) and to each other.

As VGP is almost entirely unknown, even within the scientific study of autonomic physiology, we doubt that study participants were subject to demand characteristics. It is unlikely any participant had prior cues or expectations regarding how their ability might be expected to work due to its rarity. Participants also did not have access to each other’s responses. In open-field descriptions, no participant reported any aberrant or unusual phenomena, such as goosebumps on glabrous skin areas, a fast or instantaneous piloerection response (i.e., one where the sensation is quicker than the possible latency of the SNS), or an irregular dermatomal pattern. Neither were any parethesias described –itchiness, tingling, burning, and so on. Instead, the cutaneous component of VGP, its latency, and the body parts involved appear to be congruent with normal involuntary piloerection (i.e., identical to those experienced during cold, emotional elicitation, fear, etc.). No participant mentioned any injury, neurological insult, or pre-existing condition, relevant or otherwise, in their description of how VGP was possible. Overall, the responses we received paint a consistent picture of VGP as a perfectly normal and straightforward phenomenon, one which is curious, sometimes pleasant, and essentially benign.

Having established that VGP is reported as previously described, these results both confound our understanding of voluntary control over the autonomic nervous system and provide some insight into its phenomenon. Lindsley & Sassaman (1938) debated at length whether VGP was due to an innate skill or to some form of conditioning, where an original piloerective stimuli (cold, fear, etc.) acts as an earlier unconditioned stimulus. Conditioning seems unlikely to us for three primary reasons. First, two participants discovered they could perform VGP only upon receipt of the study materials. Second, many describe stumbling across VGP either purely out of idle self-experimentation, often at a young age, or sometimes in a deliberative attempt to re-create a previous experience. For an association to form, it would have to be entirely without conscious awareness. Third, if the conditioning hypothesis were true, it would result in a remarkably consistent phenomenon which centres around a deliberative point on the back of the head or neck. There is no physiological explanation for why or how localised focus or tension is the most common physical trigger for a conditioned response.

It is also difficult to integrate the above into classic observations of piloerection from animal models. Animals display piloerection as a functional social signal in both aggression and defense (as well as during cold exposure), it can be observed on localised and generalised bodily sites, and the response can be evoked by direct stimulation of a variety of sites within the brain and periphery (Blanchard & Blanchard, 1977; Maickel et al., 1967; Shaikh, Barrett & Siegel, 1987). However, in humans, we assume the ability is vestigial and observe little interindividual variation in site (except in the case of specific medical conditions such as pilomotor seizures; Loddenkemper et al., 2004).

With respect to psychological features of VGP, most participants reported that it associated with emotional states observed during involuntary piloerection, such as absorption/immersion, and awe. This suggests that VGP-capable individuals may possibly be able to voluntarily elicit these emotional states if provoking the physical experience has direct congruence with emotional elicitation. This finding is in line with research demonstrating greater emotional reactivity after artificial induction of piloerection (Fukushima & Kajimoto, 2012). Further, approximately 72% of the sample reported using their ability to produce piloerection during various aesthetic activities. Of particular interest is one of the unsolicited uses reported, namely for attention/concentration (work, study, testing situation, lecture). Given the link between attention and emotion in terms of selection, orientation, and engagement (see Yiend, 2010 for a review), these attentional uses are congruent with reports of greater absorption and awe during VGP. Using VGP ability to control attention during situations that require it is intriguing, and a viable target for future confirmatory research.

While some differences were observed at the item level of the TIPI, none survived a correction for multiple comparisons. Thus, we leave the possibility of broad personality differences for future confirmatory research and will focus on discussing the strongest and most consistent finding—participants were noticeably higher than population norms on openness to experience. Unpacking these findings at this stage is speculative, but one possibility is that open people invest more time and effort into self-exploration. Openness is the strongest personality predictor of self-reflection (Trapnell & Campbell, 1999), and indeed some participants reported the discovery of the ability through self-experimentation and deliberate attempts to re-recreate a previous experience. However, the direction of causality cannot be assumed.

The higher openness reported by our sample does not appear to be particularly driven by one aspect of openness to experience, but rather, our sample was higher in both openness and intellect. This distinction is theorised to reflect individual differences in cognitive exploration of sensory (openness) versus abstract (intellect) stimuli and situations (DeYoung, 2015), but also emotional versus cognitive engagement with sensory and abstract information (Fayn et al., 2017). More specifically, openness positively, while intellect negatively, predicts self-reported frequency of chills (Silvia & Nusbaum, 2011). Thus, differences in openness were expected. In fact, the openness scale includes items that assess tendencies for absorption (e.g., “Get deeply immersed in music,” “Enjoy the beauty of nature”) and thus one might expect a stronger relationship with openness compared to intellect; however, the VGP sample was higher on both. While circularity in measurement could be claimed for openness, there are no items within the intellect scale that have such overlap.

Future Directions

Having established that VGP can be reliably identified in multiple individuals, the immediate extension of this work is to confirm the present findings under controlled laboratory conditions, and to add to previous observations. For instance, what overlap can be observed between other sympathetically-mediated sites in the periphery –does VGP imply a parallel response in the microvasculature, HR, or skin resistance? Moreover, does it augment or antagonise sympathetic reflex responses such as startle? Finally, if VGP is an analogue of an emotional response, does it alter the nature of emotional experience to stimuli (both naturalistic and experimental)?

These questions approach constructs of interest within broader theories of emotion; for instance, constructionist theories of emotion. As an example, the Conceptual Act Theory (CAT) understands emotion from a constructionist perspective, where they are synthesised from physically perceived representations combined with an understanding of emotional categories themselves (e.g., Barrett, 2014). Interestingly, CAT proposes that emotions are born from interoceptive information (perceived from inside the body; e.g., the sensation of the heart beating or stomach turning), exteroceptive information (taste, smell, sound, etc.), and their integration with language in a ‘conceptual act.’ In this context, VGP is unusual in the sense of being able to produce both an internal sensation (most participants describe an internal process which has a distinct activity or pathway) and a cutaneous sensation (the physical experience on the skin itself). A comparison to the experience of false haptic feedback, a manipulation where hairs on the skin are raised using electrostatic charge, without any internal experience (Fukushima & Kajimoto, 2012) would be instructive.

Limitations

There are several limitations in the present study. First, as we failed to identify any participants from a large sample of local undergraduates, the investigation presented is purely descriptive by necessity—we had no opportunity to directly study the physiology of VGP. Second, it is impossible to establish how many times our initial recruitment materials or survey were viewed, and VGP is sufficiently rare that we could not identify one in a sample of several hundred participants drawn from a general population. Thus, we have no reasonable estimate of VGP base rate. Considering this, and the fact that it has never been previously outlined, we can conclude nothing more than that VGP is reasonably uncommon. Third, while the BFAS openness/intellect measure demonstrated acceptable reliability, the TIPI measure was mostly unreliable. This is not uncommon when trying to measure personality domains with two items per domain. While the TIPI has been shown to correlate strongly with broader measures of the Big Five, short measures have been shown to increase Type 1 and Type 2 error rates (Credé et al., 2012). Further, past work has questioned the validity of the openness items within the TIPI, suggesting that they fail to capture central elements of openness to experience (Hofmans, Kuppens & Allik, 2008), which may explain the inconsistency between the two different measures of openness in our sample. Fourth, self-selected participants accessed via Internet recruitment answering a survey presumably out of curiosity might be assumed to display more personal openness. To address this requires either a carefully matched sample to any remotely collected data or identifying a VGP-capable population who are not self-selected. As outlined above, this is a significant undertaking when studying a rare phenomenon.

Conclusion

This study is the first we are aware of to outline self-reported ability to produce voluntary piloerection in a sample of more than a single individual. This phenomenon was consistent between individuals, reasonably effortless, visibly distinct, and we can tentatively conclude has correlates within the personality domain of openness. Participants who self-identified as being VGP-capable were generally aware of their ability, regard it as harmless or pleasant, and often use it to modify or prolong sensory experiences. Individuals who display VGP may play an important role within the future study of emotion and emotional regulation, as the role of the ANS integrated within the physiology and experience of visceral emotions (shock, awe, being moved, fear, panic, disgust, etc.) is potentially illuminated by individuals with rare or unusual physiology.

Supplemental Information