What homeostasis leaves out: Kinorhesis, a physiological principle of transformation
Year: 2026
Type: article
Abstract: Every physiology textbook and course teaches about homeostasis as a fundamental theory or organizing principle of physiology, and the predominating emphasis in teaching physiology is on the maintenance of a ‘stable internal milieu’ as originally articulated by Claude Bernard (1885). Despite the extraordinary impact of the theory of homeostasis, there are transformative life-cycle events that cannot be explained by homeostasis. And transformative life-cycle processes, like homeostasis, are ubiquitous in nature. Organisms experience these episodes of transformation as they undergo development and reproduction. These events clearly are physiological even though they are explicitly not homeostatic. Consequently alternative language and concepts are needed to encompass these episodes of biological transformation. In two recent papers I have proposed a complementary principle termed ‘kinorhesis’ (from Greek, propel + flow) to account for the physiological processes that control and execute episodes of transformation during development and reproduction (Horseman, 2025a, 2025b). Here I intend to summarize this principle and distinguish this idea from the many critiques of homeostasis that have populated the physiological literature. In doing so I will emphasize that a theory of kinorhesis is not a critique of homeostasis but rather a complementary idea that preserves and reinforces the theory of homeostasis. Kinorhesis is a physiological principle stating that ‘organismal life cycles are conditioned upon episodes of transformation that interrupt intervals of relative stasis so as to bring about developmental and reproductive events’. These physiological processes are predictable even though they are not homeostatic. Kinorhetic processes force the organism through changes in function, metabolism, morphology and behaviour so as to accomplish development and reproduction. There have been wide-ranging discussions and arguments about exactly what homeostasis is and how it works, and here I list a few papers that I found informative and useful guides to the rest of the literature (Bechtel & Bich, 2024; 2025; Billman, 2020; Dallman, 2003; Davies, 2016; Houk, 1988; McEwen & Wingfield, 2010; Modell et al., 2015; Ramsay & Woods, 2014). Preoccupation with homeostasis is most obviously justified by the critical importance of understanding homeostasis in the practice of medicine, and I will come back to this in a bit. But first if one takes a tour through recent literature that addresses homeostasis from a general or theoretical perspective, the individual would find numerous discussions about how we might need to reframe the concept of homeostasis. These discussions have led to a proliferation of words and phrases that are offered as being more apt than homeostasis, at least in certain situations. Among these are heterostasis, allostasis, homeorhesis, dynamic homeostasis, reactive homeostasis, anticipatory homeostasis, rheostasis and poikilostasis (Selye, 1956; Sterling & Eyer, 1988; Davies, 2016; Bauman & Currie, 1980; McEwen & Wingfield, 2010; Mrosovsky, 1990). Note that all of these terms share at least one of the root syllables homeo- or -stasis, and this is because they are all concerned with the stabilizing physiological mechanisms addressed by the principle of homeostasis. The arguments centre on how much variation is too much to be called homeostasis. Is a fever or a circadian body temperature rhythm still homeostatic, or does it need another name? If homeostasis gets disturbed (stressed) over and over again, does it then need a different name? If metabolic demands undergo a big change, do we need a new word? Though some have tried it has been hard for anyone to umpire these contests, and I am not going to try to do so. I will however offer the observation that despite these discontented offerings, homeostasis is well understood as a general physiological principle. Physicians use their understanding of homeostasis almost constantly to practice medicine, and basic biology teachers are able to convey a fundamental understanding of homeostasis without needing all the varieties of homeostasis-like words. Specialists may continue to find specific language that suits their particular circumstances, but homeostasis as a general principle has been remarkably resilient. Although some physiologists have fretted over how well the teacups are arranged on the homeostasis table, there has been an elephant in the room whose presence has been tolerated but never explicitly acknowledged. That elephant is the authentically non-homeostatic physiology associated with transformative changes that organisms must undergo for reproduction and development. In my two recent papers I have made a case for recognizing a separate organizing principle of physiology termed ‘kinorhesis’ to capture the phenomenology and mechanisms of transformative physiological changes (Horseman, 2025a, 2025b). Here is a simple example of what I mean by kinorhesis: a few days ago you were a fish-like thing swimming in a pond, breathing through gills and grazing on algae. Your thyroid gland became activated and set off a sequence of major transformative changes so that now you are hopping about on land taking air via your lungs and eating flies by flinging your tongue out to catch them. Physiology has been doing a whole host of things that are not homeostasis. It has been controlling and executing a programme of profound qualitative transformations that are functional, behavioural, morphological and metabolic. And those changes have occurred at all levels of organization, from subcellular to the whole body. Metamorphosis is an episode of transformation that has many specific facets, only some of which are understood. Episodes of transformation such as metamorphosis occur via physiological processes operating in parallel with homeostasis, but they are not explainable as homeostatic stability. There are examples of metamorphosis found across most animal phyla, and metamorphosis is only one example of transformative physiological processes. The whole domain of physiology that I am calling kinorhesis encompasses the events and processes that are generally associated with reproduction and development. A useful way of thinking about kinorhesis and homeostasis is that homeostasis operates continuously at relatively short time scales to maintain quantitative stability, whereas kinorhesis operates episodically at longer time scales to execute qualitative transformative changes. Homeostasis is necessary for survival of an individual in the present time, and kinorhesis is necessary for completing the programmes of organismal life histories and thereby perpetuating the species. The transformative tasks of physiology are complementary with homeostasis. Homeostasis does not necessarily disappear during episodes of kinorhesis, but without disrupting homeostasis the most interesting aspects of life history – development from a zygote to maturity and reproducing a new generation – could not exist. Kinorhetic physiology has always lived anonymously within physiology departments, courses and textbooks, mostly in the chapters at the end of the book and the lectures late in the semester. However the transformative changes necessary for reproduction and development have remained outside the explanatory power of homeostasis. Pregnancy illustrates this notion. A pregnant female experiences many homeostatic reflexes and adaptations. Blood pressure, nutrition, renal and immune functions all undergo well-known homeostatic responses during pregnancy (Thornburg, et al., 2006). But none of the homeostatic responses to the pregnancy explains or illuminates pregnancy itself. Pregnancy, per se, is clearly not homeostatic. It is a physiological process of profound transformations, some of which reverse after parturition whereas others become permanent parts of the maternal physiology. Understanding pregnancy comprehensively requires considering both the transformative changes and the maintenance of stabilities, and the articulations between these two imperatives. The initial paper to introduce kinorhesis approaches the topic from a philosophical and theoretical point of view (Horseman, 2025a). Physiology is the study of processes, relationships and interactions that manifest the functions of organisms. These are abstract concepts that are set within physiology's theoretical framework. Theories in physiology fall into three categories. First there are many local theories that provide explanations for function in particular cells and organ systems (Starling's law, Hodgkin and Huxley's formulations, etc.). Secondly there are generalizations from empirical findings that are mostly true, but they are not formal because there are many exceptions (germ theory, law of independent assortment, allometric scaling rules, etc.). Then there are general principles (theories) that apply across levels of organization and phylogenetic lines and provide general umbrella explanations for how organisms are alive. The ‘central dogma’ of genetic encoding and homeostasis are general principles. And kinorhesis also fits as a general principle. This first paper also describes how biology is served by both physiology theories and evolutionary theories. A second paper (Horseman, 2025b) approaches kinorhesis from a biological, rather than philosophical, point of view. It reviews a wide range of biological examples from bacteria, plants and animals to illustrate the universality of kinorhesis. Examples are described showing how kinorhesis and homeostasis can coexist benignly but sometimes come into conflict. The paper also delves into kinorhetic regulatory mechanisms and discusses how mechanisms differ between homeostasis and kinorhesis. This paper also discusses the interfaces between physiology and evolution in the context of a theory of kinorhesis. A revived interest in the physiology of evolution has recently been discussed in a special issue series of papers in The Journal of Physiology (Noble & Joyner, 2024). Further studies of these implications will be fruitful. The prominent place that has been held by homeostasis needs to be understood and appreciated. Modern physiology has its historical roots in human medicine, and medicine is mostly the practical application of homeostasis. Clinical laboratory tests generally ask which physiological variables are within the homeostatic range and whether there are informative relationships between variables that deviate from homeostatic normal. Those relationships can distinguish between simple dehydration that can be fixed by providing fluids and dehydration caused by diabetes, which will require addressing the underlying hormonal disease. Disturbances to homeostasis can decompensate rapidly if left untreated. The theory of homeostasis has become a bedrock of not only human physiology but also biology in general. Bernard's original concept was that ‘stability of the internal milieu’ was a special feature of warm-blooded animals (birds and mammals/humans), allowing them to be highly mobile and to engage in complex behaviours. Before the middle of the 20th century, physiologists focused almost exclusively on mammalian systems (Cannon, 1929; 1939). By the end of World War II, concepts of control systems in animals and their analogues in engineering were formalized by Norbert Weiner and published as Cybernetics: Or Control and Communication in the Animal and the Machine (Weiner, 1948). Theoretical treatment of homeostasis and cybernetics makes it possible to describe common control mechanisms across biology and find examples of homeostasis in invertebrates, plants and microbes. In the practice of medicine maintaining and restoring homeostasis is a primary imperative. The clinical values of physiological variables that are considered homeostatic are referred to as ‘normal values’ (or normal ranges). These clinical normal values have a loose relationship to the cybernetic term ‘set-point’, which physiologists use to designate an imaginary target value that the system will return to after having been disturbed. When the values for a regulated variable move outside the normal range, homeostasis has an increasingly difficult time, and medical interventions can be needed to avoid decompensation (breakdown of homeostasis). Values for a physiological variable therefore may be in the normal homeostatic range (near the set-point), homeostatically perturbed (outside normal values but recoverable through physiological reflexes) or decompensating (abnormal and requiring immediate intervention). Kinorhesis differs from homeostasis not only because of different outcomes (metamorphosis is kinorhetic, and stable blood sugar is homeostatic) but also because of the different mechanisms employed. These mechanistic differences can be understood generally from existing literature. However much more analysis of specific physiological instances will be necessary for a full understanding. The phenomenological categorization of physiological processes as either homeostatic or kinorhetic is an interesting exercise in itself, but the mechanistic differences have practical as well as theoretical importance. The notion of ‘mechanism’ in a physiological system has two different meanings that are helpful here. The first version suggests that a mechanism can be an a priori construct that formally describes the components of the system and their relationships. Usually represented as a diagram the a priori mechanism is not anatomical or biochemical but rather represents information flow. A mechanism diagrammed in this a priori sense represents some version of the minimal elements that logic tells us are necessary to produce the outcome, which could be an autoregulatory negative feedback reflex. The second version of mechanism is an empirical product consisting of anatomical and biochemical components that do the actual work described abstractly by an a priori mechanism. Experimental physiology is guided by a priori mechanism concepts to discover the anatomical and biochemical elements. An example of this empirical type of mechanism is the osmoregulatory negative feedback system (Carmody et al., 2015), for which a partial inventory of elements includes osmosensitive peptidergic neurons in the hypothalamus, secreted antidiuretic hormone (ADH), the bloodstream, kidneys, ADH receptors on collecting tubule cell membranes, cAMP, protein kinases and membrane vesicles harbouring aquaporin. Why is all this particularity important? The reason is not only that the particularity satisfies our curiosity but also that it provides information necessary to develop pharmaceutical or other medically useful interventions. Homeostasis and kinorhesis differ as to the abstract a priori mechanisms employed by each, and these distinctions guide the identification of empirical mechanisms. The familiar mechanisms underlying homeostasis are commonly referred to as negative feedbacks. This terminology has been widely adopted across all of physiology for both research and teaching purposes. Despite this there are some ongoing discussions in the literature about whether negative feedback is exactly the correct formulation (McEwen & Wingfield, 2010; Ramsay and Woods, 2014; Bechtel & Bich, 2025). Operationally negative feedback seems like a fine phrase for many, but not all, types of regulatory mechanisms that are required for homeostasis, and much has been learned by applying this concept experimentally. Negative feedback systems are self-regulatory because the output of the system (e.g. secreted level of a hormone) is also the input to the feedback. Certain terms (set-point, sensors, controller, etc.) are shared between physiology and engineering. It is understandable that sharing terms between physiology and engineering leads to some challenges or controversies. The set-point of a device is known because the operator chooses it, whereas a set-point in physiology can be inferred only from the behaviour of the system. Similarly components like sensors, controllers and effectors are concretely known in devices but must be discovered and inferred in physiology. Although negative feedbacks are conceptually similar in engineering and physiology, the heuristic values of feedback models are different. In engineering the model is a guide to designing and manufacturing useful regulatory devices. In physiology the regulatory device is pre-existing in nature, and the model is a guide to discovering what the device is made of and how it works. Negative feedbacks, as a general type of regulatory mechanism, are common in physiological systems, but sometimes they are complex and convoluted in ways that make our general model of negative feedback seem inadequate (Ramsay & Woods, 2014; Bechtel & Bich, 2024; Bechtel & Bich, 2025). For example body temperature regulation in mammals cannot be reduced to a single negative feedback loop. Temperature regulation involves reflexes that control core temperatures in the visceral and cerebral circulation, and regional temperatures in skin and extremities. Processes controlled by these regulatory loops include everything from shivering, brown fat metabolism, regional vasoconstriction and vasodilatation, piloerection and ancient responses like and on an on one might that body temperature is either not controlled by negative feedback or controlled by interactions negative feedbacks that have through many a view of negative feedback out other mechanisms that to homeostasis. of these homeostatic mechanisms are as simple as of and that have been by evolution to and organismal needs with An example of this from general physiology course is of to is such that the full of whereas the such as to and some metabolic the and to execute basic homeostasis & 2016; type of homeostatic regulation that is not per negative feedback is by homeostasis. In this case even though stable blood levels are there are that responses to of regulation is to other that are controlled by negative feedback mechanisms. of are the primary of changes in the This by interactions between and in and and by the of hormone and & 2016; These examples illustrate that even though negative feedback is a concept for understanding homeostasis, other stabilizing mechanisms have to to the stable internal I focused on negative feedbacks in my papers (Horseman, 2025a, 2025b) to illustrate the rather point that the stable internal on stabilizing types of physiological mechanisms. negative feedbacks are a type of mechanism that is necessary for homeostasis, can we different types of mechanisms that are required for I arguments and examples showing the of feedback loops and in kinorhetic regulation (Horseman, 2025a, 2025b). feedbacks are similar to negative feedbacks, but rather than being the output variable itself, to output from the system. feedbacks are (e.g. and In physiology feedbacks are by the system some or point that of the at parturition is a familiar that the feedback that et al., 2006). feedbacks are found across all of the and they function as of homeostatic systems, allowing the organisms to through transformations into different morphological (Horseman, 2025b). control is another type of mechanism that kinorhetic For control a physiological system is as a series of different that one another in an of a bring about the with being The predictable of developmental changes are upon physiological control processes. These developmental changes generally on and of Metamorphosis and the of and are examples of controlled kinorhetic processes. control processes in are controlled by from within and outside of the cells and feedback and control mechanisms are and physiological processes that kinorhetic changes so that development and reproduction can be feedbacks homeostatic systems, and the qualitative changes that become manifest during development and reproduction. It is is homeostasis by as the only principle for It is common to that homeostasis is principle of physiology. If one the that homeostasis is to be the only organizing principle of physiology, then there are only two ways to about the physiology of transformative these processes must be from physiology because they are not homeostatic, or the of homeostasis must be so as to include events that by their nature, the of etc.). I have never a for reproduction and development from the of physiology would be However the processes of reproduction or development are from about homeostasis. And the of physiologists to almost exclusively on homeostasis has that the elephant has in the to homeostasis so that it might include transformative processes, the only examples I have found that seem to make this to and & 2020; Billman, 2020; & & These are that and in new I not need to again, as many others that these of may have their but not in the of a place for transformative physiology, this new theoretical treatment a general of explanations for developmental and reproductive processes at the time the of homeostasis. Kinorhesis addresses episodes of transformation that are required for development and reproduction during the course of a life like homeostasis, only on mechanisms that can be not on and similar Kinorhetic processes, like those of homeostasis, are to evolutionary and In for organisms to and to new or circumstances, the regulatory processes that homeostasis and kinorhesis must the most point to emphasize is that kinorhesis and homeostasis are not Homeostasis is and at least some homeostatic processes during kinorhetic Kinorhesis is and particular This is I that these can be principles of The is not for the or of information by the than be to the for the and were by the only to for helpful discussions and to the
Source: The Journal of Physiology
DOI: 10.1113/jp291068
Author: Nelson D. Horseman
Institution: University of Cincinnati
Language: English
Subfield: History and Philosophy of Science
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $5,400
OpenAlex ID: W7140040722