The Traditional Autonomic Narrative Misleads Yet Persists – A Critical Review and Proposed Alternative to Replace It

Introduction

The traditional narrative of the functions of the sympathetic and parasympathetic divisions of the autonomic nervous system (ANS) remains ubiquitous, over a century after it was first put forward by Walter B. Cannon [1]. One can find the story summarized in any number of medical textbooks, in the biomedical literature, and on numerous websites. The following example is typical:

“The two divisions of the autonomic nervous system are the sympathetic division and the parasympathetic division. The sympathetic system is associated with the fight-or-flight response, and parasympathetic activity is referred to by the epithet of rest and digest. Homeostasis is the balance between the two systems. At each target effector, dual innervation determines activity. For example, the heart receives connections from both the sympathetic and parasympathetic divisions. One causes heart rate to increase, whereas the other causes heart rate to decrease,” [2].

Other descriptions explicitly associate two autonomic divisions with catabolic and anabolic activity, respectively [3]. Comparable descriptions are provided in most, if not all, physiology texts intended for students in biomedical fields, and are echoed throughout the peer-reviewed biomedical literature (e.g. [4]).

Typically, assertions about the nature of autonomic function are made without any citations. They reflect accepted wisdom. Neither reviewers nor editors demand support for the claims. Anecdotally, over many years of asking biomedical professionals in the clinic for their understanding of these systems, not one has offered any other description.

Despite its ubiquity, the story is not demonstrated fact. Rather, it describes an early interpretation based on extrapolations from a handful of vivid examples. Evidence accumulated over the past century conclusively shows that the story is highly misleading, as autonomic neuroscientists have been emphasizing for decades.

In summarizing the effects of sympathetic vs parasympathetic nerve stimulation in over a dozen tissues, Jänig and McLachlan wrote in 1999:

“The table shows that the idea of antagonism between the parasympathetic and sympathetic nervous systems is largely a misconception. Where there are reciprocal effects on the target cells, it can usually be shown either that the systems work synergistically or that they exert their influence under different functional conditions,” [5].

These experts in the field have between them published hundreds of peer-reviewed articles on the subject, and numerous chapters and books. In a systematic review of sympathetic cardiovascular regulation another authority, Malpas, wrote, that “[h]istorically, the sympathetic nervous system (SNS) has been taught to legions of medical and science students as one side of the autonomic nervous system, presented as opposing the parasympathetic nervous system,” [6]. His review presented exhaustive evidence contradicting this caricature. And yet, fifteen years later, and more than a quarter century after Jänig and McLachlan’s critique, essentially all pedagogy still perpetuates the numerous fictions of the traditional narrative. For the situation to change, something beyond individual critiques of specific details of the story must be required.

The premises of the present work are several. First is that the traditional narrative cannot be corrected by amendment. All its major assertions are misleading, and it distorts understanding of autonomic regulation. This is not because the compelling examples on which it was based were incorrectly described, but because of the interpretive and logical leaps by which general principles were extrapolated erroneously from those examples. Second, because the initial story is so familiar and so entrenched in modern thinking, demonstrations that individual elements of it are problematic or frankly in error will not, by themselves, cause it to be abandoned. For this to occur, the flawed underlying logic from which its generalizations arose and the many informed critiques that have been leveled against it must be reviewed systematically, so as to motivate a comprehensive rejection of it. Finally, displacing the traditional narrative will require assertion of an alternative framework for organizing observations and succinctly describing autonomic function accurately, to serve as a basis for future teaching.

The traditional narrative derives its power in part because it offers simple rubrics. The sympathetic system is responsible for “fight-or-flight”. The parasympathetic system is responsible for “rest-and-digest”. The two oppose each other. This simplicity reinforces persistence despite the embedded errors. To repeat, we must identify alternative, concise descriptions that reflect the realities of autonomic function.

The value of trying to develop such alternative descriptions has, however, been questioned. In an insightful 1997 paper, Blessing assailed the very idea of an ‘autonomic nervous system’ [7]. He argued that it was improper to consider the autonomic nervous system (ANS) as a discrete system at all, asserting that the coinage reinforces the misconception that autonomic function is separable from and independent of that of the rest of the nervous system (despite Langley’s caveats on this point [8]). He proposed that the terms autonomic, sympathetic, and parasympathetic be abandoned, arguing that we should, “refer instead to visceral neurons, both afferent and efferent. These neurons can then be seen as representing one mechanism whereby the brain communicates with the bodily organs,” [7].

In contrast, Jänig argued for the utility of the terms sympathetic and parasympathetic as they were originally defined, i.e. anatomically, but cautioned against referring to sympathetic or parasympathetic “functions” because to do so, “generates misunderstandings and gives a wrong impression of how these systems work,” [9].

In recent decades, several authorities have opted to enumerate various autonomic functions while omitting the misleading generalizations of the traditional schema [9,10,11,12,13]. Yet their approach has yet to alter broadly the teaching and popular understanding of the subject.

It is important, and possible, to identify generalizations that enlighten rather than mislead. As Langley put it, “The number of facts which have accumulated on these subjects is so vast that it is imperative to try and co-ordinate them. Every day results are published on quite trivial evidence because no general scheme is borne in mind, and because there is no standard of probability. Such results only serve to obscure the subject; if they were subjected to proper scrutiny, they might become valuable evidence for a general law,” [14].

Correcting the broad misunderstanding of autonomic function is not solely a matter of academic interest. For example, the view that the sympathetic division is fundamentally a “fight-or-flight” system underpins the widespread use of drugs targeting peripheral adrenergic signaling to manage a variety of morbidities. In the United States in 2020, over 250 million prescriptions for peripherally acting adrenergic agonists and antagonists were written, comprising over 10% of the total for the 300 most prescribed medications (Table 1). Further, the view that sympathetic and parasympathetic activities are fundamentally oppositional and counterbalanced has led to the notion that the ratio of their respective mass activities, i.e. “sympathovagal balance”, provides an index of general physiological “balance.” This idea has been used in turn to support claims that therapies that measure and adjust this ratio to achieve an appropriate “sympathovagal balance” will improve health. This idea resonates broadly with the public and continues to spur massive commercial development efforts, despite the existence of convincing critiques of this idea [15].

Most importantly, our ability to explore the integrative action of autonomic function is severely limited if we think solely in terms of whether increases in mass discharge along a given pathway increases or decreases some simple endpoint. A physiologically grounded theory of autonomic function that can provide a more useful context for hypothesis-driven research is needed urgently.

The present work suggests alternative overarching descriptions to describe sympathetic and parasympathetic regulatory responsibilities, respectively, so as to productively organize and revise our thinking about them. In the interest of limiting scope, and as with the original definition of the ANS, consideration of the important and involved topic of the relationship of autonomic output to visceral afferent input is addressed only tangentially.

The novel schema proposed here is offered to stimulate constructive debate among experts in many fields, with the goal of developing a consensus alternative to the misleading caricature that now dominates. While rejecting unjustified summaries of autonomic function, we should strive to identify functional descriptions – abstractions - that are testable, that reflect the existing state of knowledge, and that distill a large and diverse body of experimental work into a manageable schema. Such an approach should make explicit its simplifications and should not discount facts that do not fit the story. Rather, the story should be revised to accommodate the facts.

History and Critique of the Existing Autonomic Schema

Origins

Many properties of visceral motor neurons were established (e.g., [16]) long before Langley proposed the term ‘autonomic’ [17]. In his 1921 monograph on the subject, Langley reviewed earlier labels, pointing out how these distorted understanding [8]. He addressed, for example, why the terms ‘involuntary’ and ‘vegetative’ were misleading and explained why finding an alternative mattered enough that it merited a new coinage, one unburdened by existing biases.

Since at least the 1700s, the motor pathways innervating smooth muscle and glands were seen as a coherent system, distinct from somatic motor innervation, which regulated visceral activity throughout the body [18]. Originally, the entire complement of motor innervation of secretory tissue and smooth muscle, including vascular smooth muscle, was referred to as sympathetic [8,18]. Langley, however, pointed to the existence gaps along the rostrocaudal neuraxis where preganglionic nerve cell bodies of this system were absent [8,14]. He proposed restricting the term sympathetic to those pathways originating in the thoracolumbar intermediolateral column (IML), and excluding from that designation pathways whose preganglionic cells were located either rostral (tectal, bulbar) or caudal (sacral) to it [14]. Since cranial and sacral pathways had a good deal in common with each other but differed from the thoracolumbar pathways in important respects, he argued that these ‘craniosacral’ pathways should be seen as distinct subdivisions of what he termed the autonomic nervous system (ANS) [8,14,17]. By 1905 he had proposed the term parasympathetic to refer to the craniosacral division of the ANS [8]. It should be emphasized that this term indicated the anatomical relationship between the preganglionic cells of each division, but one that also implied a host of functional distinctions. For example, postganglionic sympathetic signaling was known to use adrenergic transmitters (later shown to be primarily noradrenaline [19,20]), whilst parasympathetic pathways were predominantly cholinergic [8].

Whilst Langley coined the terms autonomic and parasympathetic and conducted pioneering studies on the anatomy and physiology of the ANS, it was Walter Cannon who originated and championed the narrative of the respective functions of the sympathetic and parasympathetic branches of the ANS that has been taught ever since.

Cannon first advanced this narrative in his 1915 monograph Bodily Changes in Pain, Fear, Hunger and Rage, and modified it only slightly, adding some key caveats, in his 1932 book The Wisdom of the Body [1,21]. Modern versions follow closely the arguments, evidence, logic and conclusions of the original presentation, though rarely acknowledging its provenance.

The story is now so familiar that it can be difficult to view these systems in any other way. It is therefore important to take a step back to examine how the original arguments were constructed, and to consider the key facts, assumptions, and logic embedded in the account. Coming at the subject from a very different angle of approach – a distinct pedagogy - reveals a strikingly different picture, one fundamentally at odds with the ‘canonical’ version of the story.

In the decade preceding 1915, one line of research in Cannon’s laboratory focused on the physiological effects of adrenaline, including the emotional states in which circulating adrenaline levels rose markedly [22,23]. Here Cannon began emphasized the association William McDougall had made in his book Introduction to Social Psychology [24], between the "flight instinct" and "fear emotion," and "pugnacity instinct" and "anger emotion," later distilled into the seminal rubric “fight-or-flight”. One series of studies demonstrated that the increase in blood glucose and the appearance of glucose in the urine in response to emotional excitement (emotional glycosuria) resulted from adrenaline secretion [25,26]. Another found that adrenaline increased the force of skeletal muscle contraction and reduced muscle fatigue [27]. Yet other work showed that adrenaline injections, emotional excitement, or splanchnic nerve stimulation accelerated the coagulation of the blood, the effects of the latter two dependent on intact splanchnic innervation of the adrenals [28,29,30]. Cannon recapitulated this work from his other laboratories in the 1914 paper, “The emergency function of the adrenal medulla in pain and the major emotions,” which, as the title indicates, focused on intense emotional contexts [23]. In his 1915 monograph, he then combined these ideas with known features of digestive function (e.g. [31,32]) and autonomic control to create the now familiar template [1].

Cannon’s based his narrative on a fundamental juxtaposition of states “favourable” to digestion with those “unfavourable” to it [1]. Elevated sympathetic activity and adrenaline release were known to accompany and support physical exertion, whilst vagal activation was known to be responsible for increased gastrointestinal motility and secretion. Vagal effects were suppressed via sympathetic pathways during exertional states. Thus, sympathetic activity was treated as essentially concerned with exertional states and parasympathetic activity with digestive states, thus introducing the pernicious idea that particular nerve pathways were somehow fundamentally associated with specific behavioral states. The effects of mass excitation of sympathetic and parasympathetic nerves on pupillary diameter and heart rate were used to reinforce these caricatures. To understand how Cannon’s original presentation embedded a flawed logic, a step-by-step review of his original construction is most useful. Appendix 1 (Supplementary Material) provides such a review. The approach and fundamental construction of this original presentation formed the basis of all future teaching and remains the interpretive lens through which the system is viewed.

Critique of the Traditional Narrative of Autonomic Function

Let us consider the many problems of logic and interpretation embedded in the traditional autonomic narrative.

Inappropriate Extrapolation from Early Examples

In 1915, the first descriptions of the effects of adrenal medullary extracts [33,34], the discovery of the first hormone [35], the coinage of the term hormone itself [36], as well as the terms autonomic and parasympathetic [14] were each within one to two decades old. Thus, the functional narrative attached to these labels formed early in their history. It suffers critically from inappropriate extrapolations from early vivid examples to general rules. The problem can be appreciated more easily by considering an analogy.

Imagine that every textbook of anatomy, kinesiology or exercise physiology introduced limb function by describing the respective motions of arms and legs during typical walking, pointing out the reciprocal, rhythmic pattern of limb movements, and emphasizing that this reciprocal pattern helped maintain balance. Imagine further that this point was reinforced by examples of similar reciprocal motions during running and during swimming. The pattern holds in each case.

But what if these examples were then extrapolated to the assertion of a general principle that the arms and the legs have a fundamentally oppositional relationship, one essential to maintaining balance? This would not only be incorrect, it would actively distort understanding of the functions of the limbs, respectively, and of their relationships to each other. It would not account for the relative patterns of activity (or lack thereof) while cycling, or while playing the violin. If such counterexamples were treated as minor exceptions to a general rule formed from the canonical example of motion during walking and running, and if counterexamples were ignored by all but a community of experts, the result would be widespread misunderstanding. Arms and legs do not fundamentally exist to counterbalance each other’s actions. They have fundamentally different purposes. Sometimes they move reciprocally to help maintain balance, but it is perfectly possible to maintain balance while walking without this reciprocity, for example with the hands clasped behind the back. Any accurate description of the respective functions of the limbs would emphasize not the reciprocity seen in canonical examples, but rather, their respective roles across a wide variety of disparate tasks.

Of course, where our limbs are concerned there is no danger of misunderstanding. We know what arms and legs do. We would consider it ridiculous to assert that reciprocal motion was a general rule and was required to maintain balance. However, in the case of physiological functions that are obscure, difficult to measure and, until recently, impossible to monitor continuously and at high resolution, we overlook the logical error. The initial, convincing story based on vivid examples continues to bias our thinking, even in the face of a mass of evidence to the contrary.

The Effects of Mass Excitation Do Not Define the Regulatory Functions of a Nerve Supply

Several problems with the traditional picture of ANS function relate to aspects of nerve function that are obvious when we think about somatic motor function, but which are underappreciated when considering autonomic function. Typically, we do not speak of the overall “level” of somatic motor nerve activity. It is obvious that for somatic motor function what matters is the spatial and temporal pattern of activation, capable of producing a vast repertoire of highly articulated movements and forces, each appropriate to meeting specific goals. Further, somatic motor control is not less concerned with coordinating fine movements requiring minimal force than it is with producing powerful contractions of large muscles. The total amount of somatic impulse activity per se does not reflect the nature of the coordination accomplished by such activity. Similar levels of mass activity in a somatic nerve can occur during very different types of movements, depending on the patterning.

Yet even though autonomic impulse activity exhibits clearly organized patterns of discharge [37,38,39,40,41,42,43,44,45,46,47,48,49,50], it is frequently described solely in terms of its level - e.g. high, or low, increasing or decreasing - in the same way that we speak of hormonal signaling, as though fine spatial and temporal patterning were not the essential feature and purpose of neural regulation. As one example, the function of the sympathetic cardiovascular innervation is not to increase or decrease blood supply per se any more than the function of the somatic motor supply is to increase or decrease muscle activity per se. Rather, it is to coordinate the appropriate distribution of cardiac output within the systemic circuit to meet the metabolic needs of all tissues at all times, just as the function of the somatic motor supply is to coordinate the pattern of motions and forces needed to accomplish a wide variety of tasks. Appendix 2 (Supplementary Material) explores in detail the reasons for the tendency to view autonomic activity solely in terms of its level while ignoring the relevance of spatiotemporal patterning.

It is also critical to keep in mind that the function of a nerve supply is not defined by the effects of excitation per se. We understand the function of photoreceptors to be the detection of light (or, more broadly, the mediation of vision) and are not troubled that photoreceptors signal exposure to light via a reduction, not an increase, in neurotransmitter release. We do not claim that photoreceptors exist to detect darkness because they are depolarized in the absence of light and hyperpolarized in its presence.

Tonically active motor neurons can raise or lower the level of the activity they regulate by adjusting ongoing discharge rate either up or down. Whether the regulated variable is directly or inversely proportional to the firing rate, i.e. the polarity of the response, is a separate question, and one not fundamentally tied to the physiological factors driving variations in activity. To repeat, it is a mistake to conflate the function of a nerve supply with the effects of excitation per se.

Note, however, that this is precisely how the traditional autonomic narrative describes the roles of sympathetic and parasympathetic nerve supplies. It is common, for example, to read that the sympathetic system increases heart rate, while the parasympathetic system decreases it. Indeed, the earlier cited quote from Cannon exemplifies this flawed logic, and is oft-repeated in modern treatments: “Thus the cranial supply to the eye contracts the pupil, the sympathetic dilates it; the cranial slows the heart, the sympathetic accelerates it; the sacral contracts the lower part of the large intestine, the sympathetic relaxes it; the sacral relaxes the exit from the bladder, the sympathetic contracts it,” [1].

Similar logic is illustrated by Langley’s (1903) point that, “In the cases in which this double nerve supply exists, the nerves coming from one system do not necessarily produce a different effect from that produced by the other; but, if the effect is different, then all the central nerve-strands of one system have one effect, and all the central nerve-strands of the other system have one and a different effect.”

Such usage suggests that it is the ** nerve strands ** per se, rather than specifically an increase in impulse activity in the fibers within them, that produces the effect. The distinction may seem subtle at first glance, but it is essential, and profound. It is at the crux of the misunderstanding about autonomic function. In their review of the ANS, Jänig and McLachlan produced a table listing the effects of stimulating the sympathetic vs. parasympathetic nerves, respectively, supplying various tissues. The table’s title specified that the effects listed were effects of mass excitation [5,12]. However, most presentations are not this careful and instead imply that one or the other nerve supply, per se, is responsible for changes of a given polarity. This is misleading.

The great majority of autonomic fibers are tonically active (e.g. [12]), and changes in activity along a tonically active pathway can alter the degree of activity in the target tissue in either direction, depending on the polarity of the change. Consider for example the variation in heart rate due to respiratory sinus arrhythmia (RSA). During each respiratory cycle, variations in parasympathetic drive alternately accelerate and decelerate the heart during inspiration and expiration, respectively [51,52,53,54,55]. Similar arguments apply to pupillary control. Either branch can widen or narrow the pupil, depending on the polarity of the change in discharge. The important question concerns with which aspects of pupillary control each is concerned. (See Appendix 3 of the Supplementary Material for further detail).

The impacts of conflating the effects of excitation of a nerve with the purpose of that nerve supply are insidious and undermine a proper understanding of what these pathways regulate. When describing the functions of the sympathetic and parasympathetic innervation of a given tissue, respectively, we should give careful thought to the regulatory purposes of each, rather than default to the notion that the divisions exert generically opposing influences to achieve some “balance” simply because mass activation of the respective nerves produces opposing effects on a measured endpoint.

The case of pupillary control is instructive. The classical story suggests that the sympathetic system is responsible for pupillary dilation and that this dilation is associated with alerting situations (“fight or flight”), that the parasympathetic system is responsible for pupillary constriction and thus with “resting” or “conservative” situations, and that pupillary diameter is controlled by the balance of sympathetic and parasympathetic influences. Again, this confuses the effects of increased nerve activity per se with the purpose of the innervation. Most importantly, it overlooks the question of the functional difference between, for example, an increase in pupillary diameter due to excitation of sympathetic fibers supplying the iris dilator (radial) muscle, vs. an increase in pupillary diameter resulting from a decrease in tone of the parasympathetic drive to the iris sphincter muscle.

A more informative narrative, grounded in experimental work, would address when and why each branch exerts its effects on the target tissue. In the case of pupillary control, a large body of work reviewed in a later section demonstrates marked asymmetry in the regulatory responsibilities of the two branches. Dynamic pupillary control related to luminance and gaze, i.e. to the visual task, is clearly the province of parasympathetic regulation via fibers in the third cranial nerve. In contrast, sympathetic pathways appear responsible for varying pupil size with the state of arousal, as will be discussed. [APPENDIX ??? XXX]

These arguments emphasize the importance of how we define function or endpoint when trying to understand the regulatory responsibilities of the respective pathways in tissues with dual innervation, i.e. receiving both a sympathetic and a parasympathetic supply. The way we frame the distinct regulatory responsibilities of the respective divisions can be thought of as hypotheses, which can be explored experimentally in ways that blanket statements about oppositional effects cannot.

Sympathetic and Parasympathetic Activities Are Not Generically Counterbalanced

With respect to the question of the pattern of innervation of the two extrinsic branches of the autonomic system, Langley’s view differed in subtle but important ways from most modern treatments and is worth quoting again.

“Thus the tissues receiving efferent fibres from bulbar and sacral autonomic systems have a double nerve-supply, whilst the other tissues have but one. In the cases in which this double nerve supply exists, the nerves coming from one system do not necessarily produce a different effect from that produced by the other; but, if the effect is different, then all the central nerve-strands of one system have one effect, and all the central nerve-strands of the other system have one and a different effect.” (Langley 1903).

Despite Langley’s care in making the point that the effect of excitation did not always produce opposing effects, the idea that the two systems functioned fundamentally as counterposed opposites gained traction over the following years. Cannon championed this view initially, as noted above in the quotation from his 1915 work, but later categorically rejected the notion of autonomic balance as a fundamental principle in the discussion of his 1929 work on the effects of total sympathectomy [56]. His repudiation of this idea is reviewed in detail in the Appendix 4 of the Supplementary Material.

Gibbins criticized this canard in his review of the functional organization of the ANS: “…the insightful work of Walter Cannon has been parodied relentlessly, so that autonomic pathways end up divided into two divisions, the sympathetic and parasympathetic, that struggle against each other to maintain homeostatic balance in the face of potentially fatal stress on one hand (sympathetic “flight or fight”) or idle inactivity on the other (parasympathetic “rest and digest”). Despite the wide-spread prevalence of some version of this view in the popular and scientific literature alike, it is both over-simplistic and misleading,” [10].

This is not to say that sympathetic drive and parasympathetic drive do not at times act in opposing directions in target tissues. However, to select those cases where this is the case and to then generalize them to an overarching principle of counterbalanced regulation is unjustified. Clearly, coordination of activity between pathways with convergent targets occurs continuously, as for example happens both for the dynamic variations in heart rate and pupil diameter. But this is a different matter than saying that the two systems are fundamentally counterposed and that the balance between them is of fundamental importance.

Jänig and McLachlan’s (1999) point that the systems, “the systems work synergistically or […] exert their influence under different functional conditions,” must be kept front of mind. Extensive experimental work in many tissues supports the view that the respective branches of the ANS each regulate the activity of target tissues in distinct ways, based on distinct factors. Sympathetic and parasympathetic pathways mediate qualitatively different aspects of physiological activity, just as, for example, fiscal and monetary policy impact qualitatively different aspects of economic activity. It is this point that any pedagogy on the subject should emphasize, and we should seek to understand the nature of the qualitative differences.

Not “Fight or Flight” – Not “Rest and Digest”

The Sympathetic Is Not Fundamentally a Fight-or-Flight System

Ask most people – including most biomedical professionals - what the sympathetic system does, and they will respond that it is involved in “fight-or-flight” responses. They do this for a reason. The example of how sympathetic activation mediates diverse aspects of the coping response to exigency is used to introduce the system in nearly every overview of the ANS, at every level from the most elementary to advanced texts. If people “know” only one thing about the autonomic nervous system, it tends to be this.

As already noted, this can be directly traced to Cannon’s early narrative, based on a fundamental juxtaposition of emotions that promoted or inhibited gastric motility and secretion. Let us now turn to the logical and interpretive flaws of this fundamental association.

First is the problem of treating the pattern of physiological responses evoked by emotions “unfavorable” to gastric activity as stereotypic. Appendix 5 (Supplementary Material) recounts observations made in conscious human subjects by Wolf and Wolff that demonstrate dramatically different physiological responses of the stomach to emotional states that would fall under the rubric of ‘fight-or-flight’.

Such observations demonstrate the need for caution when forming catchall categories from disparate states having only some features in common. We may shout to warn others of danger, when coaching young athletes playing sports on a large field, or to overcome excessive background noise at a party. What each of these situations has in common is the need to use greater than normal volume to communicate. But to try to create an overarching category based on the common need for volume would be a mistake and, if used to describe essential characteristics of vocal communication, would create great misunderstanding.

Next is the problem of associating the functions of the sympathetic system fundamentally with responses to what Cannon referred to collectively as “emergency” conditions [23]: participation in ‘emergency’ responses in no way defines a system as an emergency response system.

For example, the notion that adrenaline is a “stress” hormone, and the idea that its primary function is to promote the body’s response to emergency situations was extrapolated from such unsound reasoning and persists to this day. That there is profound secretion of catecholamines from the adrenal medulla and increased sympathetic activity throughout the body, and in particular to the cardiovascular system, during times of emergency or excitement is not evidence that response to emergency is its fundamental signaling purpose, any more than the fact that we shout during emergencies or that our legs are maximally active during urgent flight merits describing either the voice or the limbs as fundamentally associated with responses to urgency or threat. Nor does demonstrating that impairing the functions of the limbs or the voice has the greatest impact on responses to exigency prove such an association. Yet in the case of adrenaline and the sympathetic system, such logic has underpinned the view of their essential purpose since Cannon’s time. The existence of careful work demonstrating the continuous regulation of adrenaline secretion as part of ongoing homeostatic regulation [57,58] has not yet markedly affected its reputation as fundamentally a “stress hormone.”

By insistently emphasizing examples of sympathetic signaling during excitement or threat, we encourage and reinforce a foundational logical leap and error. The legs are maximally active during vigorous activity – fight, flight, and any number of forms of physical exertion that are not related to either fight or flight (play, sex, etc). Paralyzing or amputating the legs impairs the most extreme forms of physical exertion most dramatically. An amputee is still able to move about a room and to maintain balance, but she is unable to run (without prostheses). However, none of this proves that the fundamental role of the legs is to engage in fight-or-flight responses, nor does it argue against their essential role as part of a general system for locomotion and balance. And were these observations used as such proof, the result would be a fundamental misunderstanding.

In the case of the autonomic system such logical errors persist because, unlike the legs or the voice, we are unconscious of the pattern of sympathetic activity during our daily lives. Appendix 6 (Supplementar Material) explores in further detail fallacies in the arguments supporting the view that sympathetic innervation is inherently an “emergency response system.”

Extensive evidence accumulated over the decades demonstrates that sympathetic activity continuously regulates the state of the internal milieu, across all behavioral states. For example, it, “plays a key role in the moment-to-moment regulation of cardiovascular function at all levels from quiet resting to extreme stimuli” [6]. This includes control of arterial pressure, cardiac output and total peripheral resistance over both short and long time spans, as well as the articulated pattern of arteriolar resistances that determine the distribution of cardiac output among the tissues, and of venular and venous capacitances that influence venous return [6,59,60,61]. Sympathetic nerves regulate the output of adrenaline and noradrenaline from the adrenal medulla in response to even slight variations in blood glucose and blood pressure, respectively [58]. The sympathetic division provides the sole innervation of the kidney [62,63], and is responsible regulating renal function including continuous influences on renin secretion, urine production and salt balance, renal blood flow and glomerular filtration [45,64,65,66]. The sympathetic division regulates various aspects of normal digestive processes including but not limited to the control of intestinal fluid fluxes, as well as regulating changes in cardiovascular function associated with digestive function [11,13,67,68]. It provides the sole innervation of adipose tissue [69,70,71,72,73] and is involved in regulating diverse aspects of adipose tissue function including mobilization and deposition of lipid reserves [70,74,75,76]. It provides the sole innervation of piloerector muscles, eccrine sweat glands, and the cutaneous vasculature and regulates thermal control continuously [77]. Sympathetic activity regulates the sleep-wake cycle via pathways connecting the hypothalamus, via the superior cervical ganglion, to the pineal gland [78,79], whose synthesis of melatonin is regulated by norepinephrine [80]. The forgoing list is not exhaustive.

Unfortunately, the incessant use of the marked physiological responses to duress as characteristic of sympathetic function generally skews our thinking about its purposes, leading to profound misunderstanding.

We also do violence to our understanding of the purpose of a given nerve supply when we think solely in terms of the effects of great increases in mass discharge, as mentioned above. The purpose of a nerve supply is to allow a spatially and temporally articulated pattern of signaling. As with speech, it is pattern, and not volume or rate, that determines specific meaning. We shout in exigent circumstances, but not only then. And our voice is not more an alerting system and less a communication system because it is more forcefully engaged during shouting than when speaking at normal volume, nor is its role in communication less important when we whisper than when we shout. It is not an apt first approximation to the truth to say that our voice is fundamentally a fight-or-flight communication system, nor that the arms and legs are fight-or-flight limbs nor the heart a fight-or-flight pump nor the sympathetic nerves fight-or-flight nerves because each is engaged at maximum intensity during emergencies. Such descriptions are not ‘simplifications’, they are frank, radically misleading errors.

The Parasympathetic System Should Not Be Described as a Rest-and-Digest System

The association of the sympathetic system with “fight-or-flight” responses is frequently counterposed to a supposed “rest-and-digest” parasympathetic sphere of responsibility. As with the “fight-or-flight” caricature of sympathetic function, this view of parasympathetically-mediated functions obscures evidence to the contrary and actively promotes misunderstanding. The examples that Cannon offered to support this view of parasympathetic function were not robust proofs. Extensive evidence accumulated since his early proposals indicates that this description also should be abandoned.

As noted, during times of duress and exertion sympathetic activity often suppresses many aspects of digestive activity and shunts blood from the abdominal viscera to the skeletal muscles to support exertional activity. But it is not the case that parasympathetic activity mediates redistribution of metabolic resources after an emergency passes, as is often implied or explicitly stated.

It is worth asking, for example, what experimental work shows that parasympathetic activity is responsible for the redistribution of blood flow from skeletal muscle to the splanchnic beds following arousal? I am unfamiliar with such work, and the absence of a parasympathetic innervation of either the skeletal muscle vasculature or most of the splanchnic vasculature (exceptions include areas within the portal circulation [81] and specific secretory tissues) suggests that the neurally-mediated component of the redistribution of cardiac output following arousal is under sympathetic control. This is not to ignore autoregulatory effects on the distribution of blood flow as visceral tissues become activated due to parasympathetic excitation, but it is important to note that there is abundant evidence of sympathetic regulation of systemic and gastrointestinal blood flow during normal digestive processes [67,82,83,84,85].

The notion that parasympathetic regulation of GI activity is only associated with digestive and restful states is also contradicted by the observation that hostile emotions powerfully stimulate gastric motor and secretary activity via vagal pathways, as has already been discussed [86,87,88]. Excitation of vagal pathways is also responsible for the marked increase in gastrointestinal motility under conditions of cold restraint stress, among others [89]. Thus, it is inappropriate to consider parasympathetic regulation as increasing gastric activity only under “conservative” or restful conditions.

Appendix 7 (Supplemental Material) reviews similar reasoning and evidence to reject the view that parasympathetic cardiac and pupillary control is primarily ‘conservative’ in nature. Such regulation is concerned not with generically opposing sympathetic excitatory influences but rather with modulating target function in response to different physiological factors than those with which sympathetic regulation is concerned.

In general, to emphasize those cases in which parasympathetic activity participates in ‘conservative’ regulatory function while downplaying or ignoring the many other aspects of parasympathetic anatomy and physiology that do not align with the rest-and-digest rubric distorts understanding of parasympathetic regulation across many tissues. Observations that are either not accounted for, or are accounted for poorly, by the traditional rubric include the roles of parasympathetic nerves in regulating aspects of sexual function, vision, respiration, and so on, and in regulating vascular function in specific vascular beds (cerebral, pulmonary, genital, etc) though not in the general systemic circuit.

For example, while there exists extensive evidence of an important role for sympathetic regulation of the sleep-wake cycle via adrenergic control of pineal melatonin production [80,90], evidence for parasympathetic control of melatonin secretion is largely lacking (despite the presence of a relatively sparse parasympathetic innervation of the pineal gland [91]). Further, several observations argue against a generalized role for the parasympathetic system in anabolism generally. For example, there is no parasympathetic innervation of either adipose tissue [70,72,73] or skeletal muscle. Further, while sympathetic fibers innervate the Leydig cells of mammalian testes and both alpha- and beta-adrenergic regulation of testicular androgen production has been demonstrated, I have been unable to locate evidence of either parasympathetic innervation of Leydig cells or regulation of androgen secretion by them [92,93,94,95,96,97].

Thus, we ought to reject ‘rest-and-digest’ as a summary of parasympathetic function, and we should not take, a priori, the presence of parasympathetic innervation of any given tissue as suggesting a “conservative” regulatory function. To do so is rather like choosing to teach that the function of the hands is primarily related to musicianship based on examples of their role in playing the piano, committing a musical score to paper, and tuning a guitar.

Distinct Spheres of Regulatory Responsibility Between the Divisions of the ANS

As noted, we can confidently reject the hypothesis that the sympathetic and parasympathetic activities are generically counterposed and reciprocal. The divisions of the ANS clearly are responsible for qualitatively different aspects of physiological regulation. The challenge is to summarize these in ways that enlighten rather than mislead and clarify rather than obscure patterns of integrative action. Ideally, such descriptions should provide a guide for exploring the interactions between the branches in situations where their respective roles are intimately intertwined, difficult to disentangle, as for example in the case of cardiorespiratory regulation.

To construct an alternative narrative, let us approach the subject from a very different starting point than usual. Rather than beginning with the effects of mass activation or blockade of one or the other branch in tissues with convergent innervation from both, let us begin by considering situations in which one or the other system is wholly or largely absent. From there, we can proceed to situations where a predominant or exclusive role of one or the other branch in specific regulatory activities is well-established. This approach will help clarify the responsibilities of each branch and aid in developing hypotheses about the roles of each in tissues where their respective actions are more difficult to isolate.

Sympathetically Regulated Functions

As already noted, the sympathetic division provides the sole innervation of the kidneys, as well as of the eccrine sweat glands, piloerector muscles, adipose tissue, and the bulk of the systemic vasculature (with specific exceptions, to be discussed later). What does this pattern suggest regarding broad sympathetic regulatory responsibilities?

Let us begin with the renal innervation. The kidneys plays a critical role in controlling the internal milieu. They receives a rich, tonically active, sympathetic supply [38,40,41,98,99]. Surprisingly, very little attention has been given to the absence of renal parasympathetic innervation [62,63]. The implications of this fact for the interpretation of the respective responsibilities of the two branches of the ANS has not received much attention either. I have yet to find an introductory text that lists the kidney among the list of tissues receiving only sympathetic, and no parasympathetic, innervation. Many introductory texts, in fact, continue to prominently feature either Netter’s classic 1977 illustration of the pattern of whole body autonomic innervation ([100], Section IV Plate 2, p. 70) or derivatives of it, that include a parasympathetic pathway innervating the kidney [100].

In their 1983 HRP retrograde tracing study, Norvell and Anderson demonstrated the absence of renally-projecting neurons in any parasympathetic nucleus (DMV, nucleus ambiguus, sacral segments S1-S3) [62]. In motivating their study, they cited prior work that had observed acetylcholinesterase positive renal nerve fibers but which had not characterized these fibers nor described their origins. They noted that Barajas et al. had already provided evidence that the renal fibers and terminals expressing acetylcholinesterase were sympathetic [101,102]. It is possible that the bias toward treating dual innervation as a rule caused and continues to cause their demonstration of the absence of renal parasympathetic innervation to be overlooked or dismissed. As of April 2022, PubMed listed only seven citations of Norvell and Anderson’s HRP study. Scopus listed twenty-two such citations, several of which have received broader attention. One of the citing works was a more recent confirmation of their finding of an absence of any parasympathetic renal innervation using a different method - fluorogold labeling via microcapsules applied to the renal plexus [63]. The lack of parasympathetic renal innervation has important implications for understanding the respective regulatory responsibilities of the branches of the ANS and should be central to the teaching of the subject.

The kidney controls long-term blood volume and blood pressure [98,103], hematocrit, whole body fluid and salt balance, and it plays a central role – in cooperation with the lungs – in controlling the pH and buffering capacity of the extracellular fluids. Introductory texts, if they mention sympathetic control of kidney function at all, tend to focus on the sympathetic innervation of the juxtaglomerular apparatus (JGA) and its influences on renin secretion, particularly in the context of hypertension. In fact, all areas of the nephron and its vasculature are sympathetically innervated, and thus all renal functions –glomerular filtration, regulation of plasma electrolyte and pH balance, osmolarity, urea cycling, etc - are subject to central control via sympathetic pathways [66,98,104]. Differential sympathetic control of distinct aspects of renal function, based on activity pattern and level, and population of fibers activated has been demonstrated [40,42,45,66,98,105].

Cardiovascular regulation via the sympathetic innervation of the kidneys and systemic vasculature occurs continuously [6,38,40,64,106]. Sympathetic pathways control circulating blood volume via renal regulation of salt and water balance, via renal control of hematocrit, and via control of venular capacitance (particularly in the splanchnic beds) [66,67]. By controlling the arteriolar resistances of parallel pathways in the systemic circulation, the sympathetic system regulates the proportion of cardiac output that flows through each organ, and along parallel pathways within each of organ. Sympathetic adjustments in venular capacitance are regulated by a subset of sympathetic ganglion cells distinct from those responsible for regulating arteriolar resistance, demonstrating the articulated nature of such control [107]. By adjusting cutaneous blood flow, piloerection, brown adipose tissue thermogenesis and eccrine sweating, the sympathetic plays key roles in thermoregulation [77]. As the sole motor nerve supply to adipose tissue, sympathetic innervation regulates fat depot mass, cellularity, and lipolysis [70,108].

Taken together, these observations indicate that the CNS regulates via sympathetic pathways the distribution of metabolic resources and maintains the appropriate physico-chemical environment to support current and anticipated action of the tissues, throughout the body and at all times. As Cannon himself noted in The Wisdom of the Body, “The sacral and cranial divisions of the interofective system, however, operate only indirectly and somewhat remotely to assure a constant state. It is the middle or thoraco-lumbar division which acts promptly and directly to prevent serious changes of the internal environment,” ([21], Chapter XVI, p. 262).

Here, it is important to emphasize again that changes in sympathetic activity can adjust the level of regulated physiological variables either up or down, depending on the polarity of the change in activity (i.e. whether it is waxing or waning) [109,110,111,112,113]. This is one reason that antagonists have effects opposite to those of agonists. Additionally, distinct adrenoceptor subtypes mediate distinct and in many cases opposing effects [114,115]. Sympathetic co-transmitters afford an additional mechanism to differentially modulate sympathetic regulation. A counterbalancing parasympathetic influence is not required for the highly articulated regulation of innervated targets. Further, sympathetic regulation of physiological variables is not limited to “emergency” situations. It is not unitary (up or down) but highly articulated, and it occurs continuously, across all physiological states.

This includes digestive states [68]. The digestive hormone, CCK - released by intestinal enteroendocrine cells in proportion to the luminal concentrations of protein and fat - mediates increased blood flow in the upper GI tract via inhibition of specific sympathetic vasoconstrictor pathways [82,85,116]. Sympathetic pathways drive changes in cardiac output, heart rate, and aortic pressure in anticipation of feeding [68,83,84,117]. Sympathetic regulation of gut blood flow occurs as part of the digestive process and not solely or even primarily under conditions of extreme excitement or physical exertion [67] and the sympathetic system plays a dominant role in regulating intestinal fluid fluxes [11].

The types of sympathetic regulation described above are necessarily integrated across all the tissues of the body. The central nervous system makes coordinated, continuous, temporally and spatially articulated adjustments to physiological activity everywhere in the body via sympathetic pathways. These pathways allow anticipatory control over the global distribution of metabolic resources, and the ability to incorporate experience and learning into optimizing solutions to complex, dynamic and multidimensional physiological challenges.

Despite continuous sympathetic regulation of physiologic state, Cannon et al. showed clearly that the organism can survive complete sympathectomy [56]. How should this be interpreted? As Blessing emphasized, it is the nervous system as a whole, and not any subset of pathways comprising it, that ultimately regulates the physiological state of the body [7]. Thus, in the extraordinary situation of an absence of sympathetic pathways, the brain and body may compensate via other means to maintain physiological variables within the ranges necessary to maintain the integrity of the organism, though presumably without the full regulatory flexibility and resultant physiological “virtuosity.” This issue will be addressed in greater detail in a later section.

To summarize then, in contrast to the emphasis on sympathetic function under conditions of outward exertion - 'fight or flight’ responses – a more accurate description would emphasize that the sympathetic division generally regulates the internal milieu, continuously, under all physiological states, irrespective of whether anabolism or catabolism dominate.

Parasympathetically-Regulated Functions

As already noted, the “rest-and-digest” description of parasympathetic function is as misleading as the fight-or-flight caricature of the sympathetic system. If so, then how might we summarize and generalize parasympathetic spheres of regulatory responsibilities?

Proposed Rubric for Understanding Parasympathetic Functions

Despite his emphasis on “rest-and-digest” activities, Cannon made another observation about the craniosacral innervation in passing that provides a useful jumping off point for thinking about parasympathetic function. Of the craniosacral (i.e. parasympathetic) division he wrote:

“Furthermore, the two divisions – sacral and cranial – are similar in being largely subject to interference by the movement of striated muscle. Just as contraction of the bladder and rectum can be aided or checked by nerve impulses from the cerebral cortex, the reactions of the pupil to light or to distance can be induced by voluntary acts. Indeed, as a rule, the workings of the sacral and cranial divisions involve the cooperation of the cerebro-spinal nervous systems to a much greater degree than do the workings of the sympathetic division, because they are much concerned with external orifices surrounded by striated muscle,” [21].

Cannon’s remarks are prescient, though some caveats are in order. Cannon’s suggestion that striated muscular activity “interferes” with parasympathetic control of structures connected with the orifices of the body is unfortunate. It can more aptly be described as coordination between somatic and parasympathetic motor functions toward a common purpose. It is also possible to take issue with the characterization of the cooperation between the sympathetic and “cerebro-spinal” systems as being lesser, quantitatively, than that between the craniosacral and cerebrospinal systems. The former relationship is qualitatively different from the latter.

These concerns aside, viewing parasympathetic activity as regulating smooth muscle and secretory activity associated with dynamic interactions with the outside world via the “external orifices surrounded by striated muscle” - involving intimate coordination with specific patterns of skeletal muscle function - and with movements of the head more generally, allows for a less biased exploration of the common features of craniosacral regulation. It has the great advantage that it encompasses many roles of the parasympathetic innervation that are excluded from the “rest-and-digest” rubric, while avoiding biases that the rhyme perpetuates.

For example, parasympathetic activity clearly regulates adjustments in lens curvature with changes in the fixation point via contraction of the ciliary muscle [118]. Loss of parasympathetic, but not sympathetic, innervation impairs this function. Parasympathetic regulation of lens curvature is coordinated in concert with somatic regulation of the ocular (skeletal) muscles controlling eye movements. Together, these coordinated ocular functions allow for effective tracking of and focus on visual targets. Thus, the parasympathetic supply to the ciliary muscle controls a basic visual function that has nothing to do with either rest or digestion or anabolism, and it is not markedly counterbalanced by sympathetic influences [119]. Lens accommodation is intimately coordinated with the activity of the ocular muscles, motions of the head, and with visual function per se. It is not concerned with vision only during resting or conservative states. We will discuss this point further when considering control of pupillary diameter, which is regulated by both parasympathetic and sympathetic activity.

As with control of the curvature of the lens, in both the bladder and the colorectum, voiding is under parasympathetic control [120]. Voiding requires coordination with the activity of the skeletal musculature, including the activity of the external urethral and anal sphincters, respectively, in addition to changes in the activity of the muscles of the trunk [121]. Obviously, as with eating, conscious behaviors are involved in urination or defecation as well. While relaxation of internal sphincters during voiding is parasympathetically controlled, in a healthy individual these changes follow a decision to void, changes in position associated with the act, and specific patterns of activity in the abdominal and trunk musculature. Sympathetic denervation or blockade affects the temporal profile of voiding, but the effect is minor and temporary, while parasympathetic regulation is required for normal function [121]. The nature of the counter-regulatory relationships between sympathetic and parasympathetic control of bladder tone will be addressed in a later section on tissues with dual innervation.

In the upper GI tract, skeletal muscles bring food and drink into the mouth, prepare it by mastication and initiate swallowing. Depending on the species, the esophagus is invested with a greater or lesser degree of striated muscle rostrally. Skeletal muscular activity in the mouth, pharynx, and upper esophagus is coordinated with parasympathetically-regulated secretory activity and smooth muscle activity in the more distal esophagus during swallowing, as well as with relaxation of the lower esophageal sphincter and receptive relaxation of the gastric reservoir [31,122,123,124]. In the case of the genitals, reflex erection and vaginal lubrication – i.e. that resulting from external stimuli - are mediated by the parasympathetic system, although interestingly, and in contrast, psychogenic erection and vaginal lubrication result from changes in sympathetic activity [125].

Taken together, these observations suggest that diverse parasympathetic functions might be aptly described as coordinating specific patterns of smooth muscle and secretory activity involved in interactions of and exchanges between cranial or pelvic structures and the outside world. Of course, such exchanges may also require changes in the distribution of metabolic resources, e.g. in the distribution of cardiac output, and so it would be wrong to claim that the sympathetic division is uninvolved. But the nature of sympathetic involvement is qualitatively different, and not as directly tied to the specific patterns of skeletal muscle activity as is typical of parasympathetic activity.

Considering parasympathetic regulatory responsibilities in this way also dovetails nicely with the proximity of somatic motor nuclei to parasympathetic motor nuclei. As the topic is presently taught, students can be forgiven their surprise when they first learn that the vagus contains a population of somatic motor fibers arising from the nucleus ambiguus (which fibers are responsible for control of the laryngeal muscles). Parasympathetic and somatic fibers also comingle in the oculomotor, facial, and glossopharyngeal nerves, and the somatic and parasympathetic nerve nuclei supplying these fibers are closely apposed to each other [126,127]. Similarly, the sacral intermediate gray essentially merges with the ventral horn, while in thoracolumbar segments the intermediolateral column is well-separated from the ventral horn [126,127].

Of course, close coordination with skeletal muscle activity need not be seen as a fundamental feature of all parasympathetic regulation. It is doubtful that such coordination is its exclusive purpose. The description offered is not intended to exclude functions not necessarily coordinated with specific patterns of skeletal muscle activity, e.g. hepatic or pancreatic secretions related to the digestive process.

We will return to the question of how the relationship between thoracolumbar and craniosacral innervation in tissues receiving convergent innervation might be interpreted, with special attention given to the innervation of the bladder, gut, heart and iris - canonical examples used in the traditional narrative to illustrate a supposed fundamental counterbalancing oppositional relationship between the activities and functions of the two branches.

First, however, let us consider autonomic function in the airways and the pulmonary vasculature, topics excluded from the traditional autonomic pedagogy.

Parasympathetic Innervation of the Airways and Pulmonary Vasculature

Airways

Extensive parasympathetic innervation of airway smooth muscle occurs in all air-breathing vertebrate species examined [128,129]. Vagal preganglionic fibers innervate intrinsic ganglia located around the airways, and their postganglionic fibers innervate airway smooth muscle and mucous glands [130,131].

At least two distinct types of vagal motor fibers regulate airway diameter, and thus resistance. Cholinergic parasympathetic drive to airway smooth muscle is bronchoconstrictive, while nitrergic parasympathetic drive is bronchodilatory [132,133,134]. The cholinergic pathway is known to be tonically active. Decreases or increases in the level of cholinergic drive increase or decrease airway diameters, respectively [131,135,136]. Thus, the parasympathetic innervation does not cause bronchoconstriction only; rather, it regulates airway diameter, increasing and decreasing conductance in an articulated pattern as needed to meet physiological demands, about which more later. Various stimuli – hypercapnia, hypoxemia, hypotension, airway irritation – elicit bronchoconstrictive reflexes via vagal efferent excitation, while lung inflation causes bronchodilation via bronchodilatory reflexes that inhibit vagal efferent firing [128].

In contrast to an extensive parasympathetic supply, sympathetic innervation of airway smooth muscle is much sparser in most species (excepting the guinea pig, in which it is somewhat more well-developed), while in man it is said to be nearly or completely absent [130,131,132,137,138]. However, airway smooth muscle in all species examined expresses high levels of β₂-adrenergic receptors, which modulate airway diameter in response to variations in the level of circulating adrenaline [139,140,141]. Thus, while overall airway resistance - and thus total ventilation as a function of the atmospheric-alveolar pressure gradient – is indirectly sympathetically-modulated (via adrenal medullary adrenaline secretion), the neuroanatomy in many species including man argues against a direct role for the sympathetic division in fine spatiotemporal control of airflow within the respiratory tree. The physiological significance of the parasympathetic supply of the airways has long been a subject of conjecture [142,143] but the innervation pattern indicates its capacity to regulate the pattern of airflow along parallel bronchiolar pathways.

Due to the clinical significance of asthma and other disorders of the airways, the neurochemistry, anatomy, and physiological effects of stimulating or inhibiting discrete subpopulations of airway parasympathetic fibers has been studied extensively. However, as Otis wrote in 1983, “[a]lthough we now have a considerable understanding of the mechanics of bronchial smooth muscle and of the effects of numerous physiological and pharmacological agents on its behavior, the exact role it plays in normal physiological function is unclear. Numerous plausible suggestions have been made, but none has been convincingly demonstrated.” [142]. Gabella echoed the point over a decade later [143]. Uncertainty over the physiological function of this nerve supply persists to this day, despite steady progress in understanding detailed mechanisms controlling airway conductance (cf [133]). Still, the conservation of this parasympathetic supply to the airways throughout phylogeny suggests that it must be of significant adaptive value, especially since its dysregulation can lead to debilitating and potentially lethal respiratory pathology [144]. The ongoing uncertainty concerning the physiological role of this nerve supply highlights the importance of a theoretical context for investigating it.

Pulmonary Vasculature

Another important fact almost universally excluded from autonomic pedagogy concerns the substantial parasympathetic (vagal) innervation of the pulmonary vasculature [145,146,147,148]. This contrasts with the lack of parasympathetic innervation in most of the systemic vasculature. The extent of parasympathetic innervation of the pulmonary vasculature varies between species, but it is notably less dense than the sympathetic innervation of the same vessels [146,148,149,150]. There remains disagreement concerning the degree, but not the presence, of parasympathetic innervation of the human pulmonary vasculature [146,148,151]. The functional significance of this innervation, like that of the parasympathetic innervation of the airways, remains obscure. This is an interesting situation, worthy of consideration.

Functional Implications of the Parasympathetic Pulmonary Innervation

The lungs mediate gas exchange between the atmosphere and the circulation. Respiration requires ventilation - the delivery of fresh, oxygen-rich air to and expiration of CO₂-laden air from the alveoli. Gas exchange takes place via diffusion in the alveolar capillaries, which must be adequately perfused. Alveolar ventilation and perfusion vary dynamically throughout the lungs. Efficient gas exchange requires local matched ventilation and perfusion throughout the respiratory tree [152]. Local ventilation-perfusion mismatches reduce the efficiency of whole-lung gas exchange even when ventilation and perfusion are matched globally, resulting in “physiologic dead space”, i.e. regions of the respiratory tree where gas exchange with the pulmonary capillaries is incomplete due to V/Q mismatches [152,153]. Elaborate autoregulatory mechanisms exist to adjust respiratory epithelial blood flow with alveolar airflow and vice versa. Research into V/Q matching has focused on adjustments to various imposed manipulations or conditions but typically has not addressed pre-action (i.e. feedforward regulation) based on expectation or learning.

Little is known of the dynamic coordinated articulation of ventilation and perfusion throughout the lung across a range of physiological conditions. Few studies have inquired into a possible role for the parasympathetic innervation of the airways and blood vessels of the lung in matching ventilation and perfusion to achieve optimal blood gas exchange and minimize physiologic dead space. Still, it is of interest that Allen et al. noted, that human nerve fibers considered to be parasympathetic on the basis of vasoactive intestinal polypeptide (VIP) immunoreactivity, “seemed most abundant in the small muscular arteries just proximal to the respiratory unit – arteries that can be considered as the resistance vessels of the lung. Extensive innervation of these vessels suggests that the flow of blood into the respiratory unit and capillary bed is partly under neurogenic control,” [150]. Thus, the anatomical sites at which these fibers are reliably found at greatest density are appropriate to regulating perfusion of the respiratory epithelium.

Our ability to study the dynamic regulation of ventilation-perfusion is limited. In a behaving animal it is a daunting technical challenge to monitor the fine structure of rapidly and regionally varying pressures and volumes – comprising a combination of cyclic and non-cyclic components - and to analyze the impact of these changes on V/Q variations. But the ability to coordinate complex spatiotemporal dynamics is one of the defining features of a nervous system, and so it merits our attention.

Added to the inherent difficulty of studying the problem is uncertainty about those conditions under which such regulation would meaningfully affect performance. To draw the analogy with Cannon’s findings on the effects of total sympathectomy in animals kept in the laboratory [56], without a specific hypothesis about the purpose of a particular motor nerve supply, and given multiple reinforcing mechanisms to secure the most vital physiological functions, determining the function of that nerve supply using only ablation studies and/or global manipulation and measurement of key variables may be difficult, if not impossible. For example, various authors have cited Fritts et al (1958) observation that acetylcholine infusions into the pulmonary artery did not markedly affect gas-exchange performance as evidence for a minor parasympathetic role in pulmonary hemodynamic regulation [148]. However, infusing a bolus of a neurotransmitter into a large vessel does not and cannot mimic the coordinating action of a nerve supply delivering spatiotemporally patterned signaling in a dynamic situation, particularly one demanding optimal performance.

Thus, just as we should not take Cannon’s finding - that sympathectomized animals kept in the laboratory maintain various physiological variables within a tolerable range - to imply that the function of sympathetic innervation is primarily to mediate “emergency” responses, so also we should not take the difficulty of demonstrating the regulatory functions of the parasympathetic supply to airways and pulmonary vessels under normal circumstances to mean that such innervation is irrelevant to ongoing physiological regulation. In fact, it would require the assertion of such an ongoing role and a belief that it must be physiologically important to justify the research effort needed to explore it in a meaningful way.

Irrespective of whether the parasympathetic innervation of the airways regulates V/Q matching in physiologically important ways, what is certain is that these two fundamental functions of the lung – regulation of the distribution of airflow and regulation of the distribution of blood flow, respectively - are each subject to articulated parasympathetic control. Further, the sparse to absent sympathetic bronchiolar innervation in most species including man indicates that adrenergic signaling in these species cannot coordinate an articulated pattern of alveolar ventilation.

Respiration involves exchanges of gases between the external environment and the body, via the nose and mouth. The pattern of pressure changes responsible for these exchanges is determined by the action of skeletal muscles. Intrathoracic pressures vary continuously with postural and positional changes, and with the respiratory cycle, and these ongoing physical variations impact both airflow and perfusion in complex ways throughout the thoracic cavity. Vagal motor drive to both the airways and the blood vessels is tonically active and is subject to modulation via vagal afferent input from a variety of receptor types, including J receptors, pulmonary stretch receptors, and others [154]. Regulation of airway resistance is important during exercise. For example, Hesser and Lind concluded that, “neural mechanisms compensating for internal flow-resistive loading play an important role in the control of ventilation during exercise, both at normal and at raised air pressures,” [155]. During exercise, there is an increase in the rate at which both respiratory movements and movements of the head and trunk occur, and the demand for efficient respiratory function is maximal. These and other observations, taken together, suggest that parasympathetic control of the pulmonary system may be quite important under “fight-or-flight” conditions, that its pattern is tightly articulated with specific patterns of skeletal muscle activity, and that such regulation may also occur across all behavioral states. Further, while the physiological significance of the parasympathetic regulation of airway resistance and pulmonary perfusion are not yet established, whatever their significance, it is highly unlikely to be relevant only during anabolic or ‘restful’ states.

Extending the Hypothesis to Relationships in Tissues with Extensive Dual Innervation

General Considerations

As noted earlier, in most descriptions of ANS function, the regulatory purpose of the innervation is usually erroneously identified as equivalent to the effects produced by shocking whole nerve branches or superfusing neurotransmitter receptor agonists onto target tissues. This leads to formulations such as, “The autonomic nervous system is important for homeostasis because its two divisions compete at the target effector. The balance of homeostasis is attributable to the competing inputs from the sympathetic and parasympathetic divisions (dual innervation).” [156]. Again, the polarity of response to adrenergic vs. cholinergic agonists is not informative as to physiological function.

It was proposed above that the thoracolumbar division be described as responsible for regulating the physico-chemical environment of the tissues, and the distribution of metabolic resources appropriate to present and anticipated demand, at all times. A possible terse summary of this is that the sympathetic division serves as the body’s quartermaster, and/or the “regulator” of the physico-chemical state of the body.

Rather than describing parasympathetic function in terms of a specious essential complementarity to sympathetically-regulated functions, it was then proposed that craniosacral function, broadly speaking, be described as concerned with dynamic interactions and exchanges with the outside world via the orifices of the body (cranial, and sacral) - often requiring close coordination with specific patterns of skeletal muscle activity. As will be described below, it clearly also is concerned with coordinating visceral activity with specific movements of the head and trunk, e.g. adjusting systemic blood pressure in response to rapid positional changes, or heart rate with the respiratory cycle, and possibly also with a similar role regulating cerebral blood flow. A possible summary description for the parasympathetic division, then, is that it serves as the “coordinator” of interactions with and exchanges between the internal and external milieux.

Still, each division has responsibilities that could be considered to fall under the proposed rubric of the other. For example, sympathetic regulation is involved in differentially distributing (coordinating) blood flow related to specific patterns of expected muscular activity [110]. Conversely, parasympathetic (and skeletal muscle) effects on gas exchange obviously have a significant role in regulating the pH and buffering capacity of the internal milieu. Still, if the sense in which these terms are intended is understood, they may serve as useful rubrics.

By describing sympathetic and parasympathetic pathways as having qualitatively different regulatory responsibilities, and by calling attention to what is known about these differences, the door is opened to more productive interpretations of the relationships between them in tissues innervated by both. This yields more physiologically meaningful questions than the traditional view of a stereotypic antagonism. The interactions between the branches depend on the tasks being accomplished and are not consistent, but rather, are highly variable, and dependent on the situation. In the case of autonomic regulation of the heart specifically, Kollai and co-workers have demonstrated this directly, via simultaneous recordings of the cardiac sympathetic and parasympathetic nerves under a variety of conditions [157,158,159,160,161,162,163,164,165,166], as will be discussed below.

With respect to the concept of “autonomic tone”, it is certainly the case that in tissues receiving a dual innervation, the activity of both branches influences the overall activity of the tissue. However, where dual innervation occurs, rather than seeing the functional relationships as analogous to those between flexor and extensor muscles, as originally proposed and later repudiated by Cannon [1,56], perhaps a better analogy is the relationship between pyramidal (corticospinal) and extrapyramidal (vestibulospinal, rubrospinal, reticulospinal, tectospinal) tracts innervating the spinal motor pool. Each of these pathways mediates distinct aspects of motor control. Pyramidal pathways are described as broadly concerned with directed movements, while the various extrapyramidal pathways are described as regulating muscle tone, balance, posture, and locomotion. When considering pyramidal and extrapyramidal inputs to a lower motor neuron, the default assumption is not that they exist to counterbalance each other. Thus, their respective roles can be investigated without such bias. Here, Saper’s comment about ANS premotor areas that “the interconnectivity of its components, located at virtually every level of the neuraxis, is more similar to a network than a strict hierarchy,” is relevant [167]. The output systems and the reflexes that modify them are unlikely to be organized for the purpose of producing a reciprocally balanced pattern of impulse activity in the two divisions per se.

Below, the proposed rubrics of quartermaster and coordinator, respectively, are applied to suggest interpretations of the distinct regulatory responsibilities of each branch in the canonical examples supposedly establishing a fundamental sympathovagal antagonism: the bladder, the GI tract, the heart, and the iris/pupil.

The approach here is not exhaustive nor could it be. It is offered to illustrate how these propositions about the regulatory responsibilities of each branch can lead to novel, testable hypotheses regarding function. Where the physiological functions being regulated by the respective branches are intimately intertwined, we should expect complex relationships between them, both peripherally and centrally. Whether the relationships are oppositional or not will depend on the specific circumstances.

The Bladder

It is helpful to begin with the bladder since its physiological functions are straightforward, and the neurobiology of bladder control has been studied extensively [120,168].

The functional repertoire of the bladder is limited to two roles: storage and elimination. The bladder stores urine produced by the kidneys until such time as a decision is made to void, at which point it switches from a storage mode to a voiding mode. Urine production is continuously regulated by the renal sympathetic nerves [66]. Accommodation of the bladder moderates increases in intravesicular pressure resulting from the added volume. Bladder accommodation is primarily under sympathetic control [168], and thus the control of accommodation can be seen as related to and coordinated with the sympathetic regulation of urine production by the kidney, which is ongoing.

In contrast, the episodic decision by the organism to urinate is accompanied by the movement of the body to the appropriate place and posture for urination. Voiding involves relaxation of the external urethral sphincter via withdrawal of ongoing tone in the somatic motor supply to it, and movements of the skeletal muscle of the pelvic floor. In coordination with these articulated changes in somatic motor activity, the detrusor muscle is contracted via excitation of parasympathetic cholinergic fibers, increasing urethral pressure. Concomitantly, relaxation of the internal urethral sphincter is stimulated by increased activity in parasympathetic nitrergic fibers [120,168]. Thus, bladder voiding is controlled by the parasympathetic innervation acting in close coordination with the somatic motor system.

Transitions between the storage mode and voiding mode of the bladder are described as “switch-like” and the neural circuits controlling the lower urinary tract exhibit this switch-like behavior [120]. In the storage mode, parasympathetic influences on detrusor tone are inhibited by sympathetic inputs, and control of internal urethral sphincter tone is modulated by adrenergic fibers as part of a ‘guarding reflex’ which prevents inappropriate urination. During urination, parasympathetic output uncouples the sympathetic drive from the detrusor muscle, allowing the parasympathetic system to coordinate contraction of the bladder.

The storage function of the bladder may be seen as related to the rate of urine production and thus to regulation of whole-body fluid and salt balance. Conversely, the voiding function involves coordinating smooth muscle activity of the bladder and internal sphincter with skeletal muscle activity in mediating exchanges with the outside world (i.e. micturition) episodically, and when appropriate. As these two functions - relaxation in concert with filling and contraction during voiding - are physically opposed processes, the systems are arranged so that one or the other dominates, depending on circumstance.

GI Tract

In the bladder, filling is associated with the sympathetically-regulated production of urine by the kidney, and thus with regulation of the internal environment, while emptying is the result of conscious decisions to void. In the upper GI tract, in contrast, the situation is reversed. Filling of the stomach results from conscious decisions and the actions of skeletal muscles required to eat and drink. Vagal parasympathetic outflow coordinates smooth muscle and secretory activity with the movements of skeletal muscles involved in chewing and swallowing. This coordination allows for the lubrication and passage of food to and through the lower esophagus, relaxation of the lower esophageal sphincter, and receptive relaxation of the gastric reservoir [31,122,169,170,171]. Thus, filling of the stomach, unlike filling of the bladder, involves parasympathetically-mediated coordination with the skeletal musculature during episodic exchanges with the outside world. Increased vagal cholinergic drive is responsible for stimulating gastric acid secretion and motility during the cephalic phase of digestion [172]. Vago-vagal reflexes, mediated by vagal afferent feedback, modulate the pattern of gastric acid secretions and motility. It is worth emphasizing that the great majority of enterogastric vago-vagal reflexes are inhibitory [173], that is, they result in a reduction in vagal cholinergic drive to the stomach (and nitrergic drive to the pylorus) and a reduction in motility and acid secretion, not its stimulation (see for example [174]). Thus, it is inappropriate to see parasympathetic control of the gut as solely stimulatory.

Emptying of the gastric contents into the duodenum occurs through and is controlled by the motility of the antropyloric region of the stomach. Antropyloric coordination allows articulated control of the particle size distribution and composition of chyme that passes into the duodenum [175,176]. The vagus nerve coordinates both increases and decreases in gastric emptying, depending on circumstance. Increased vagal nitrergic drive to the pylorus and increased cholinergic drive to the antral region serve to propel contents into the duodenum by relaxing the pylorus and contracting the antrum, respectively, while vago-vagal reflexes evoked by a variety of duodenal stimuli inhibit both [174,175]. Such regulation is intimately related to immediate downstream digestive and absorptive capacity. However, the pylorus is also richly invested with sympathetic fibers, whose excitation results in increased pyloric pressure and reduced gastric emptying [177,178]. Given that both divisions of the ANS can and do restrict gastric emptying, we may ask, under which conditions are the activities of one or the other the more significant factor? This question is considered in detail in Appendix 8 (Supplementary Material).

Analyzing the situations in which one or the other extrinsic branch can be shown to exert dominant effects on various GI functions may reveal generalizations of both conceptual and therapeutic value. Viewing the sympathetic and the parasympathetic divisions as responsible for different aspects of regulation provides a basis for exploring the great variety of patterns of interaction between them. The existing caricature of a generic opposition does not.

The Heart

In the traditional pedagogy, the contrasting effects of sympathetic vs. parasympathetic excitation on heart rate are presented as a canonical example of a fundamental juxtaposition between the branches. A typical formulation: “For example, the heart receives connections from both the sympathetic and parasympathetic divisions. One causes heart rate to increase, whereas the other causes heart rate to decrease,” [179]. As noted earlier, the notion that one or the other branch adjusts activity in only one direction, and the related but distinct misconception that the function of a nerve supply is identified only by the effects of increases in discharge, are both erroneous.

In exploring the relationships between cardiac sympathetic and parasympathetic outflows, we have the advantage of direct electrophysiological evidence. In an extended series of investigations, Kollai and co-workers recorded activity in the cardiac sympathetic and vagal (parasympathetic) nerves simultaneously, during a variety of experimental manipulations [157,158,159,160,161,162,163,164,165,166,180].

These studies are summarized and analyzed in Appendix 9 (Supplementary Material). They demonstrate clearly that a wide range of phase and amplitude relationships between sympathetic and parasympathetic cardiac outflow may occur, depending on the specific physiological challenges being met. Thus, as noted with respect to the error of extrapolating from a vivid example – e.g. the reciprocal motions of arms and legs during typical walking or running - to a general rule, the clear “reciprocal” pattern of activity observed in some states does not justify the assertion of a general rule regarding functional relationships between the branches supplying the heart.

This brings us to the question, then, of how to interpret the relative roles of each branch in cardiac regulation. If the two branches regulate qualitatively different aspects of cardiac function, which of course must be coordinated, the question arises, what are these? Let us consider the regulatory challenges of cardiac control.

Heart rate and stroke volume, and thus cardiac output, vary continuously. Over any given interval, enough blood must be pumped by the heart to meet the perfusion demands of all tissues at pressures appropriate for each. At the same time, cardiac activity must be coordinated with rapid fluctuations in intrathoracic pressure and position in the gravitational field (each of which influence preload, among many other factors) resulting from the activity of the skeletal muscles controlling respiratory and postural movements. It must also be coordinated with changes in preload and afterload resulting from changes in sympathetic regulation of the capacitance and resistance vessels, respectively [107].

Changes in ventilation, in position, and in the activity of the ‘muscle pump’ all vary the rate of venous return continuously, leading to mismatches in right and left ventricular filling and output that, since the system is a closed loop, must be balanced over time, but which are not matched instantaneously [181,182,183]. Guz et al. found that cardiac pacing in dogs increased swings in left and right stroke volume with the respiratory cycle, indicating that respiratory sinus arrythmia (RSA) may modulate such variations [181], at least in this paradigm. Dynamic changes in venous return due to changes in ventilation, position, and the activity of the muscle pump are not directly related to the total systemic demand for cardiac output. These factors vary much more rapidly than does systemic demand for cardiac output. Given a common total demand but a difference in movement and position within the gravitational field and/or the respiratory cycle, cardiac activity will differ.

The proposed alternative narrative suggests that we view cardiac sympathetic regulation as concerned with adjustments of cardiac function related to and integrated with changes in sympathetically-mediated vasoconstrictor drive to resistance and capacitance vessels, and with changes in systemic demand for cardiac output generally. In contrast, parasympathetic drive is seen as concerned with cardiorespiratory coordination and coordination with postural/positional changes, i.e. with the rapid variations in intrathoracic pressure and pressure gradients associated with skeletal muscular activity and movement. These two types of regulation necessarily impact each other. For example, it is clear that parasympathetically-mediated variations in heart rate have effects on arterial pressure waves at the respiratory and slower frequencies (e.g. Mayer waves) and on sympathetic rhythmicity [15,184]. It is therefore unsurprising that the two systems would show a high degree of central convergence [185,186,187,188,189].

Appendix 10 (Supplementary Material) reviews in detail the evidence and arguments concerning such a ‘division of labor’ in cardiovascular autonomic control. The key assertion of the proposed schema is that it is the parasympathetic innervation that is most immediately responsive to the rapid changes in intrathoracic pressure and body position, which themselves are determined by musculoskeletal and somatic motor activities, while the sympathetic cardiac innervation is predominantly concerned with systemic demand for cardiac output.

It may be argued that the formulation proposed here for describing the respective regulatory roles of sympathetic and parasympathetic innervation of the heart is too simplified and too speculative to serve as the basis for introductory teaching of the subject. However, introducing the established facts concerning the functional asymmetry and the non-stereotypic relationship between sympathetic and parasympathetic influences first, followed by the caveat that the extrapolated theory is just that, a theory, seems preferable to perpetuating the caricature that the sympathetic innervation increases heart rate whilst the parasympathetic system lowers it, which is clearly false but remains the basis of present teaching. The opinions and critique of experts in this field will be needed to refine the proposed description to a point that it can be widely accepted as a replacement.

The proposed alternative schema promotes the formulation of testable hypotheses, and there is value in doing so. Viewing each branch as responsible for distinct aspects of cardiac function is likely to shed light on the significance of observed heart rate variability in various physiological contexts [6,15,190]. Critically, if we view the two divisions as regulating different things, then we need not think in terms of a generally “optimal” level or pattern of heart rate variability (HRV), a metric that has been used to estimate the ratio of the mass activity in the two motor pathways supplying the pacemaker. Rather, the level of activity in each branch, and the need to vary heart rate, depends on the specifics of the physiological situation [191]. For any given type of physical activity certain consistent patterns of relationship between mass discharge in the two branches may be discernible and/or demonstrably optimal. But there is no reason to suppose that the ratio of the magnitudes of mass activity in the two branches supplying the pacemaker (or those of any other nerves) must reflect some generalized global state of optimality any more than, say, the ratio of total movement of the arms to the legs might be considered optimal. Such ratios may be informative in certain situations – e.g. evaluating performance during bicycling, swimming, tennis, pitching or batting – but it would be surprising to find that a particular ratio was optimal across all physical activities, or that it served as some sort of broad measure of physiological balance or well-being. Specific hypotheses about why the level of activity in each branch changes as it does in various situations may help us understand better the information content in records of heart rate variability and thus make much more effective clinical use of HRV data [192,193,194].

The Eye

Autonomic control of the pupil is another of the canonical examples used to illustrate a supposed fundamentally reciprocal relationship between parasympathetic and sympathetic activity. Yet, as with the systems already reviewed, there is ample evidence of marked differences in the regulatory responsibilities of ocular sympathetic and parasympathetic innervation.

The classical view of the parasympathetic system as responsible, generically, for pupillary constriction and the sympathetic system as responsible, generically, for pupillary dilation is unjustified. Rather, as reviewed in detail in Appendix 12 (Supplementary Material), abundant evidence supports the view that the two branches regulate distinct aspects of pupillary function. Variation in the activity of either branch may widen or narrow of the pupil, the response depending on whether tonic activity is waxing or waning or vice versa.

Based on the available evidence (Appendix 12) the parasympathetic supply can be described as coordinating the activity of the ciliary and pupillary muscles in concert with the demands of the changing visual field, including changes in luminance and fixation point. The activity of the ciliary muscle and iris smooth muscles must be integrated with that of the skeletal muscles controlling vergence, saccadic, and other eye movements [195,196]. In contrast, the sympathetic supply appears to have the dominant responsibility for regulating pupillary responses associated with various types of ‘arousal’ [197], although this term is so broad, covering so many different conditions, that care should be taken to better define this category. This claim regarding arousal, though asserted forcefully by Loewenfeld [198], may be an oversimplification because several studies indicate an important role for central inhibition of parasympathetic drive – sometimes referred to as central sympathetic inhibition due to its adrenergic nature. The functional antagonism between sympathetically-dominant dilator muscle activity and parasympathetically-dominant sphincter activity can be thought of as resembling the similar functional antagonism between accommodation and voiding in the bladder, though in the pupil a much wider array of phase relationships in the activities must occur, due to the essentially infinite potential combinations of the need to respond to the visual field and to arousal of various sorts.

Application of the Hypothesis to Other Tissues

Asserting a theory of the general functional spheres of responsibility of the two divisions of the ANS provides a lens through which to consider their potential respective functions in specific tissues where we remain ignorant of these. Taking such a position on functional responsibilities leads to the formation of testable hypotheses that are more enlightening than statements about the polarity of response to excitation of one nerve or another.

For example, for the endocrine pancreas the proposed framework suggests that the parasympathetic innervation controls hormonal secretion related to absorption and storage functions, while the sympathetic innervation is most relevant to the need to match metabolic energy supply to anticipated demand more broadly. Similarly, the hepatic parasympathetic supply may be concerned with regulating postabsorptive adjustments in glucose and lipid storage, while the sympathetic innervation may be responsible for glucose and fatty acid regulation related to expected systemic demand for energy, both in concert with the corresponding regulation of the activity of the pancreatic islets.

As another example, for those vascular beds that receive parasympathetic innervation – in contrast to the bulk of the systemic circulation that does not – the proposed framework suggests that while the sympathetic innervation is concerned with matching and adjusting metabolic supply and demand, the parasympathetic innervation is concerned with what might be broadly referred to as hydraulic factors, such as the need to supply fluid volume for exocrine secretion, engorgement of the genitalia, or considerations related to the need to maintain constant intracranial volume.

Appendix 12 (Supplementary Material) addresses these ideas in detail and reviews available evidence supporting them.

It should be emphasized that to explore such hypotheses, it is not enough to demonstrate that stimulation of a given nerve supply produces a given effect. One must determine when such effects are relevant. This is difficult. Even in cases where the goal is limited to identifying the therapeutic potential of mass activation (for example by an implanted stimulator device), understanding the regulatory significance of the activity is important for interpreting and addressing potential sequelae of any given stimulation regime.

Experiments designed to probe hypotheses regarding each branch’s sphere of regulatory responsibility can promote important conceptual advances. Since it is typically much more difficult to discern clear functional responsibilities than it is to demonstrate the effects of changes in mass activity on endpoints, situating experimental work in the context of a broader theory can help justify the required effort and funding.

Anticipation and Dynamic Regulation

Many physiological studies and most teaching focus attention on how physiological responses to imposed perturbations or manipulations act to restore some physiological variable to a range that is considered normal and stable, i.e. toward some resting ‘set-point’. This emphasis is also traceable to Cannon, whose coinage of the term homeostasis was intended to elaborate on and improve upon the then-dominant principle of the constancy of the internal milieu that Claude Bernard had asserted half a century before [199]. Cannon proposed his neologism to account for the variation in physiological parameters in the ‘resting state’, and to draw attention to the mechanisms called into play both to maintain certain variables within a “normal range” during action as well as those that act to restore the resting state after the organism has acted. As previously noted, Cannon et al. had demonstrated that the organism can survive total sympathectomy [56]. This begs the question, why do we need sympathetic nerves at all? Put another way, why does the organism commit vast resources to maintaining this network, if it can survive in its absence?

Autonomic innervation allows the organism to adjust the activity of innervated tissues not only in response to imposed changes, but in anticipation of need, i.e. before an error signal is present. It permits integrative activity with a high degree of spatial and temporal patterning. Feed-forward mechanisms allow for experience and learning to improve performance. The importance of anticipatory change and the resulting fluidity of action led Sterling and Eyer, in the 1980’s, to coin a new term, allostasis, which they proposed as a superior basis for understanding physiological regulation [200]. The principle of allostasis emphasizes the fluidity of physiological change, how feedforward signaling and positive feedback facilitate such fluidity, and the role of anticipation in physiological regulation, in contrast to the homeostatic model, which emphasizes the ‘dynamic’ stability of physiological variables and mechanisms of negative feedback.

That the autonomic system mediates physiological changes in anticipation of action was well-established since before the work of Pavlov on conditioning over 120 years ago [201]. Notions of “feed-forward” or central command-driven adjustments in cardiorespiratory function, for example, have a long history. Waldrop et al. trace it back at least as far as the late 19^th century [202], citing Johansson’s hypothesis that a central command mechanism could account for the rapid circulatory changes accompanying onset of exercise [203,204]. Later studies demonstrated that stimulation of discrete central loci can elicit coordinated behavioral respons[203,204es, along with physiological changes that anticipate demand [110,205]. Eckberg (2003) described the long history of research into heart rate variations associated with respiratory sinus arrythmia (RSA). These were shown by Frédéricq [206] to persist in the absence of associated intrathoracic pressure changes (i.e. during open chest surgery), and by Heymans [207] to persist after pulmonary denervation. Such findings suggested, and more recent studies confirm, that respiratory-associated variations in parasympathetic drive to the cardiac pacemaker are initiated by a central pattern generator, whose output is modified by peripheral input from baroreceptors and other sources [208].

For the ENS, where our knowledge of the pattern of connectivity and the properties of the network elements is extensive, modeling of the coordination of peristaltic activity in isolated gut has illustrated the value and flexibility of feedforward regulation [209]. There is no reason to suppose that this finding is unique to the intrinsic network of the gut.

Challenges to Discerning the Nature and Significance of Autonomic Function

It is somewhat remarkable that despite decades of interest and study, we still lack explanations for the physiological importance of the prominent parasympathetic nerve supplies of the pulmonary and cerebral vasculature and that of airway smooth muscle, of widely studied parasympathetically-mediated phenomena such as RSA, of many of the varied prominent frequency components apparent in sympathetic nerve activity, of the sympathetic innervation of the thyroid [210], and many other notable features of autonomic innervation. Why is this?

Two reasons stand out. One has to do with limitations in the spatiotemporal resolution with which we can monitor and discern dynamic changes in physiological activity of various sorts under ‘normal’ conditions. The other concerns functional redundancy.

In reviewing cerebral blood flow (CBF) regulation, Claassen et al. addressed both these points. Concerning the latter point, they wrote,

“An important characteristic of CBF regulation is mechanistic redundancy, i.e., overlapping mechanisms contribute to maintaining CBF under highly challenging conditions. Studies exploring the regulation of CBF are importantly impacted by this, because the overlap in pathways makes it difficult to explore the relative importance of individual pathways or identify key contributors. From a teleological perspective, this redundancy makes the regulation of CBF a robust system where multiple strategies are present to ensure precise control and thus protect against potential brain damage” [211].

Distributed regulation is difficult if not impossible to monitor effectively without adequate spatial and temporal resolution. Indeed, development of higher-resolution methods has repeatedly opened new vistas of inquiry and understanding. Thus, Claassen et al. pointed out that it was only with the advent of high-resolution methods for measuring rapid changes in CBF that a dynamic component of autoregulation was finally discerned [211]. It may require the development of the ability to measure blood pressure and flow simultaneously with broad spatial and high temporal resolution in a behaving subject to allow us to determine whether the cerebral parasympathetic vascular supply (PVS) adjusts vascular reactivity with arterial pressure changes resulting from anticipated voluntary movements of the head and trunk, as suggested in Appendix 12 (Supplementary Material). The same holds true for the proposals made concerning the regulatory roles of the pulmonary PVS and the parasympathetic innervation of the airways in ventilation-perfusion matching throughout the respiratory tree with high spatial resolution under dynamic conditions, where it would also be necessary to monitor pressure and airflow in the airways simultaneously.

This begs the question of the physiological importance of mechanisms that, to be discerned, require such elaborate and sensitive methods, especially if other mechanisms can substitute in their absence to achieve organismal goals. Are they then superfluous? The problem is similar to that presented by Cannon’s findings on the effects of total sympathectomy in animals maintained in the lab, and his interpretation regarding the significance of sympathetic function generally [56].

Again, the key point concerns the value of anticipating imminent change, pre-action, and learning over time how to optimize function, even as the “equipment” changes with age and state. Negative feedback requires an error signal to be corrected, and involves a time lag for its correction. Biological systems have multiple reinforcing mechanisms for achieving negative feedback locally, but feedforward systems that allow for adjustments in advance of an anticipated change can reduce or eliminate errors before they occur. This permits a level of flexibility and fluidity not possible otherwise, as pointed out by Sterling and Eyer in their discussion of allostasis [200]. Autonomic innervation allows for feedforward regulation, based on learning and experience.

But can we recognize its importance? We easily recognize the importance of learned pre-action in producing the fluidity of trained movements, such as handwriting and artistic, musical or athletic performance. We also immediately can notice impairments of somatic motor control based on sometimes subtle, sometimes glaring disruptions of ‘normal’ motion in those afflicted. And we know that those afflicted with motor disorders can perform many desired tasks at a level that could not be distinguished from ‘normal’ by simple measurements. Various basal-ganglia related movement disorders can be distinguished from each other and from cerebellar disorders by characteristic movement patterns, some of which are subtle and require medical training and specific tests to notice. Yet measurements of the time afflicted subjects might take to move from one point to another, or of their ability to cook a meal, or play a musical instrument, could easily fail to detect any impairment at all. Conversely, imagine two musicians both with unimpaired motor capacity, one professional, another an amateur of average skill, each playing a sequence of notes. What quantitative measure of performance could be used to distinguish virtuosity? The number of notes played per minute? The mean duration between notes struck or the variability in these intervals? Variations in amplitude?

We recognize physical virtuosity immediately because we have everyday familiarity with the patterns it produces. It emerges from ceaseless repetition and thus, from effective anticipation of and adjustment in preparation for what is about to occur. It cannot be captured easily with simple measures. In contrast, we typically have no such familiarity with what constitutes what might be called physiological virtuosity, no way to recognize or quantify it, and so we are ignorant of its importance. We struggle to identify the pathways responsible for producing it, because we do not know how to manipulate or characterize it. But this does not mean such virtuosity or the pathways responsible for producing it are unimportant, nor that they only matter under exceptionally demanding circumstances. Indeed, feelings of vibrant good health and well-being may reflect such virtuosity. Conversely, states of malaise and many syndromes that we identify as functional disorders may reflect impaired feedforward function.

Until appropriate recording and analysis methods let us discern directly the articulated spatiotemporal patterning that an autonomic nerve supply may orchestrate, hypotheses about such regulation may elude direct investigation. And even then, many challenges may remain, as the example of the difficulty of discerning musical virtuosity by objective measures of performance makes clear. Perhaps the only system equal to the task of monitoring the effects of autonomic activity with appropriate spatiotemporal resolution is the ensemble of afferent fibers whose integrated activity likely contributes to the pleasing sensations and emotions associated with autonomic virtuosity, or the aversive ones associated with autonomic dysregulation. While we await technological developments allowing us to investigate these things experimentally, thought experiments, reasonable guesses and theory can help guide our thinking. Certainly, a nerve supply is a biologically costly apparatus, and its persistence throughout phylogeny argues strongly for its importance, even when we remain ignorant of how it is important. We should not fail to point out our blind spots, and it is helpful to be clear on the reasons for our ignorance.

Possible Objections to the Proposed Pedagogy

Any proposal that departs so markedly from tradition as that offered here is bound to be controversial, particularly given the prevailing concepts that have driven autonomic research and the interpretations of results over the past century. Indeed, it would be surprising if experts in each of the areas surveyed above did not question interpretations presented. For example, it may be objected that the description of the sympathetic sphere of responsibility does not adequately capture its role in the allostatic adjustment of setpoints of targeted processes. Other potential objections are addressed in Appendix 13 (Supplementary Material) .

Ideally, such objections, and any ensuing debate, will lead to refinements of the proposal or to suggestion of a superior alternative. The ultimate goal should be to arrive at a consensus framework that can represent robustly the opinions of the community of experts. It is hoped that, sooner rather than later, a consensus can be reached that the popular view of the sympathetic and parasympathetic divisions as generically oppositional and counterbalanced must be rejected and replaced in toto. Ideally, discussions of alternative hypotheses may stimulate research interest in and support of basic studies designed to address these concepts. It is difficult in the present environment to get such work funded. This kind of work is painstaking, and its implications will frequently be open to further debate. Often, experimental work will simply reveal greater complexity, and such complexity is not always welcomed.

Still, consider the perhaps billions of dollars’ worth of both public and commercial funding directed toward the measurement and analysis of cardiac sympathovagal balance, motivated by the notion that the ratio of total sympathetic to parasympathetic drive to the cardiac pacemaker (in those cases where the measurement can be shown to reflect this ratio robustly) somehow reflects the overall state of physiological well-being, i.e. a balance between arousal and repose. This specious notion – still widely propagated - should be understood to have been founded on a speculative story originally proposed over a hundred years ago by Walter Cannon, based on relatively sparse data, and later rejected by him [1,56]. If we understand sympathetic and parasympathetic drives as responsible for qualitatively different aspects of physiological regulation, then their ratio in any given nerve or globally takes on no more general a significance than any other ratio of disparate though possibly related activities, e.g. the ratio of the movements of the arms to those of the legs. Such ratios may indeed be informative in specific cases and may serve as useful proxies of state. But to confer on them broader significance, representing an overall state of balance and coordination, is a mistake.

In the effort to develop superior descriptions of the respective regulatory responsibilities of the divisions of the autonomic system, we must recognize that the nervous system did not evolve to have its various pathways conform to some overarching set of design principles, but rather to solve problems necessary for survival. Blessing’s point - that it is the nervous system as a whole rather than any component of it that is ultimately responsible for meeting physiological demands – is critical [7]. The regulation of blood gas levels illustrates his caveat about attributing to subsystems the responsibilities of the whole. Blood gas concentrations are controlled by the ventilatory movements of skeletal muscle, and thus it is an error to argue to suggest that autonomic activity controls the internal milieu independent of somatic activity. And of course, behavior plays a critical role in all physiological regulation. Nonetheless, we can agree that while the kidney, gut, and liver are all crucially involved in cardiovascular function, none are responsible for actually pumping the blood. Thus, it makes sense to analyze the activity of the cardiovascular system as an identifiable subsystem with responsibility for a specific set of tasks, so long as we do not lose track of its inextricable relationships to the whole. The same holds for the ANS.

4.1. The Proposed Alternative Narrative

The alternative schema proposed here centers known and significant facts that are typically excluded from and are inconsistent with the traditional functional narrative. Starting from those physiological phenomena whose regulation can be shown with confidence to be the sole or predominant province of one or the other division of the ANS, hypotheses concerning the qualitatively distinct spheres of regulatory responsibility of each are developed. These are then applied to the interpretation of the actions of each branch in tissues with dual innervation.

The key features of the proposed schema are as follows:

• It argues against seeing the typically opposed effects of mass cholinergic vs. adrenergic stimulation or blockade as evidence of an essential oppositional relationship between the systems. It holds that respective divisions of the ANS use different final neurotransmitters to regulate different aspects of physiological function, and that the polarity of response to mass activation of each is not indicative of the regulatory function per se.
• It develops a description of sympathetic regulatory responsibilities based on observations from tissues that only or predominantly are innervated by the sympathetic branch, including the kidney, adipose tissue, the bulk of the systemic vasculature, sweat glands, the piloerector muscles, and the adrenal medulla (the first two of which are typically overlooked). It associates sympathetic function generally with the continuous regulation of the internal milieu, as Cannon pointed out [21], across all behavioral states and not just aggressive/defensive states.
• In contrast, it describes parasympathetic regulation broadly as concerned with coordinating secretory and smooth muscle activity with interactions and exchanges with the outside world—eating, breathing, speaking, voiding, looking, mating, moving, etc—often closely articulated with associated movements of the skeletal muscles. This dovetails with the anatomical observation that nuclei containing parasympathetic preganglionic neurons are closely apposed to those of somatic motor neurons, and the physiological observation of the much lower latencies in the responses of target tissues to cholinergic compared to adrenergic stimulation. Both may be related to the demand for rapid, fine coordination of musculoskeletal activity with many, though presumably not all, parasympathetically-controlled smooth muscle and secretory activities. This view of parasympathetic regulatory responsibilities can be applied productively to form reasonable hypotheses regarding the role(s) of parasympathetic nerve supplies whose significance remains obscure, including those of the pulmonary circulation and of airway smooth muscle, or the cerebral circulation, among others.
• It advocates an emphasis on the unique capacity of autonomic outflow to mediate anticipatory adjustments in smooth muscle and secretory activity and deprecates seeing this outflow as solely responsible for responding to changed circumstances after they occur.

The proposed alternative re-orients common observations of the traditional narrative and suggests more straightforward associations in other cases. For example, the proposed alternative holds that what is similar about parasympathetically-mediated cardiac and pupillary regulation is not supposed “conservative” functions, but rather a parasympathetic role in regulating smooth muscle and secretory activity associated with interactions with the outside world in which the somatic system also participates. In contrast, it notes the frequent finding in many tissues of a more slowly varying sympathetic drive governing tonic levels of activity over which dynamic parasympathetically-mediated modulation is superimposed. For example, it sees ongoing parasympathetic tone to the heart not as an ongoing brake on cardiac energy expenditure or workload, as is sometimes claimed, but rather as centering the dynamic range of parasympathetic effects on heart rate in either direction. It sees parasympathetic regulation of the GI exocrine secretions as comparable to that governing respiratory, lacrimal and other systems, involving exchanges with the outside world and requiring increased blood supply to meet ‘hydraulic’ demands involved in such interactions rather than to meet metabolic needs, and not to any overarching “rest-and-digest” sphere of responsibility.

Interestingly, the proposed schema turns the traditional association of sympathetic activity with outward exertion and parasympathetic activity with inward regulation somewhat on its head, or perhaps more properly, on its side. That is, it frames sympathetic activity as primarily concerned with regulating the conditions of the internal milieu for both visceral and somatic tissues, and parasympathetic activity as coordinating internal smooth muscle and secretory activity with specific patterns of somatic motor activity associated with largely conscious acts as noted above, i.e. the decision to consume, to void, to mate, to breathe, to move, to look, etc. This view is consistent with the rostrocaudal distribution of thoracolumbar vs. craniosacral preganglionic nuclei. Langley’s original rationale for the term parasympathetic was anatomical, and the alternative narrative suggested here harmonizes with that distinction.

It is hoped that the proposed approach will stimulate demand for experimental work designed to clarify which aspects of regulation are the primary responsibility of each respective branch in tissues receiving dual innervation. Hopefully, the broad functional labels of quartermaster and coordinator proposed here to summarize sympathetic and parasympathetic actions, respectively, are neutral enough to avoid introducing biases that interfere with revising them in the face of experimental results to the contrary, as has been the case for the traditional narrative. The nervous system did not evolve to allow its various functions to be summarized succinctly, and it is important not to try to force facts to fit a narrative. Rather, narratives must be adjusted to encompass the facts.

The effort to replace the existing autonomic narrative is unlikely to succeed without concerted and persistent effort on the part of experts in many fields. What is needed is a consensus-building process of refinement, followed by widespread dissemination of the resulting consensus. The need is urgent.

I know of no other branch of science in which its fundamental principles and its pedagogy are so widely acknowledged by authorities to be fallacious, yet which has nonetheless doggedly resisted revision. The challenge to any change in thinking on the subject is in part the vast infrastructure of texts, study materials, test questions and so on that reinforce the current view, and in part the deep popular entrenchment of the story and the biases it has produced. Knowledge of the canonical examples and their traditional interpretations are required to demonstrate competency on professional entrance and licensing exams. Thus, being steeped in these distortions is a prerequisite for becoming a biomedical professional, and the system reproduces itself, even in the face of decades of commentary by the most knowledgeable people in the field that the story does not hold.

Teaching students that the traditional narrative is incorrect will lead to problems if, when they take entry, qualifying or licensing exams to continue in their field or seek study materials online, they are confronted with the errors of that narrative, presented as truth. Displacing the existing mass of misleading pedagogical tools and test questions will not be easy, and it cannot be accomplished by any small group of advocates. It does no good to eliminate one of these errors in a single textbook. The whole mutually reinforcing panoply of study materials and the tests they are designed to prepare for must be revised. This is a challenge that must be taken on by the field generally, and taken on in toto, if we are to correct and uproot the persistent errors and distortions that lie at the heart of current misunderstandings of autonomic function.