**Links to**: [[A humorous take on hormones]], [[Cortisol]], [[Principle]], [[Sundowning]], [[Principle of Sufficient Cortisol]], [[Neuroendocrinology]], [[Active inference]], [[08 Active ignorance]], [[Hegel]], [[Foucault]], [[Pathology]], [[The Normal and the Pathological]], [[Melencolia I]], [[Personalized medicine]], [[Secretion]], [[Motion]], [[Movement]], [[Conceptual movement]] [[Free energy principle]]. Presentation given at [Psychiatrie en Filosofie conference: The Paradoxes of Madness](https://www.psychiatrieenfilosofie.nl/too-mad-to-be-true-iii).
# _Phänomenologie des Krankengeistes_ (or: the Dialectics of the HPA-axis): A neuroendocrinologically-informed, philosophical account of allostasis under the influence of corticosteroids
 
>That’s the only way you can change the world, you can either move a muscle or secrete something.
>
>Karl Friston, interview with Neil deGrasse Tyson, 2024.
 
>So one might say: voluntary movement is marked by the absence of surprise. …“But why isn’t one surprised here?” … When people talk about the possibility of foreknowledge of the future they always forget the fact of the prediction of one’s own voluntary movements.
>
>Ludwig Wittgenstein, _Philosophical Investigations_, 1953, sections 628-9 (1986, tr. Anscombe).
 
>I would propose more forcefully that there is not in itself an _a priori_ ontological difference between a successful living form and an unsuccessful form. Moreover, can we speak of unsuccessful living forms? What lack can be disclosed in a living form as long as the nature of its obligations as a living being has not been determined?
>
>Canguilhem, _The Normal and the Pathological_, 1904.
 
>_Ni Metre, Ni Dieu._
>
>H. P. Duerr, 1974, p. 49.
<div class="page-break" style="page-break-before: always;"></div>
# _Phänomenologie des Krankengeistes_ (or: the Dialectics of the HPA-axis): A neuroendocrinologically-informed, philosophical account of allostasis under the influence of corticosteroids
**Abstract**: The parameters which determine a state of (mental) ‘health’ are debatable, if not altogether objectively impossible to determine (Fava et al. 2019, Sterling 2020). In psychosocial allostasis, these parameters are especially difficult to delineate because they pertain to the dynamic sociocommunicative frameworks in which a person is embedded: predictive bidirectional adaptations between an agent and their embedding result in the preservation of temporary ‘ideals’ of what counts as (collective) health (Schulkin & Sterling 2019); essentially what is worth _survival_. In the context of “cognitive health”, although medical typologies are continually redefined, generally, the hallmark of (temporary) “altered” or aberrant states such as psychosis has been the perception-production of delusions (Lopez-Silva & McClelland 2023, Carhart-Harris et al. 2014), which, by and large, seem to lead to what **from an outside perspective** appear as (communicatively) dysfunctional. From an active inference perspective, ‘delusions’ can be framed as errors of precision estimation, and dysfunction can thus be understood as a result of allostatic overload: predictive models become overwhelmed by stimulus and/or seek actualization in order to relieve said overload (Parr et al. 2022, Clark 2023). But **where** (in the individual or in the semantic context), exactly, the error is located, or the stress as overload stemming from, is the question. The interest in this piece is in meditating on some of these aspects, in order to highlight the relationship between individual and collective psychosocial allostasis as inextricably intertwined with the evolving **concept of *stress*** itself. In the context of patients self-administering corticosteroids: the concept of stress bypasses an autonomic reaction. Therefore, even more than for all those (i.e., _all of us_) already mediated by the concept, the semantic refinement and dialogical dynamics of what “stress” is, and its effects, come to play a central role in cognition and adaptation.
<small>Keywords: allostasis, altered states, psychoneuroendocrinology, distributed cognition, active inference, self-narration and confabulation, dialogical reasoning.</small>
<div class="page-break" style="page-break-before: always;"></div>
### Introduction
 
>The entire world is under stress, we no longer have a control group.
>
>James Herman, recipient of the *Ron de Kloet Award*, _Stress NL Symposium_, January 30 2025.^[This distinction is awarded to an international researcher making outstanding contributions to the understanding of stress in adaptation, resilience and vulnerability for disease.]
 
>... renegotiations of public metaphors which channel personal experience of the body, particularly in relation to reproduction, immune system functions, and ‘stress’ phenomena; ... struggles over meanings and means of health in environments pervaded by high technology products and processes: ... If we learn how to read these webs of power and social life, we might learn new couplings, new coalitions.
>
>Haraway, _Cyborg Manifesto_, pp. 171-2.
 
### Introduction
The biochemical etiology of allostatic overload at the level of the individual is the result of perceived interoceptive and/or environmental stress, which elicits a neuroendocrine response: the release of cortisol (McEwen 2005, p. 316). Cortisol, the end-product of the hypothalamic-pituitary-adrenal (HPA) axis, is produced in anticipation as well as in response to stress, and its vital dynamics emerge from feedforward and back-regulation between the pituitary gland and the adrenal cortex (Russell et al. 2023). Despite decades of research on the psychiatric symptoms associated with corticosteroids (Sirois 2003), showing how corticosteroid-excess and deficiency induce neuropsychological changes—even in short-duration therapy (Russell et al. 2023)—an account of how they affect the prediction of an allostatic sense of self or semantic coherence remains elusive. This has mostly been due to the fact that dysfunction of HPA-axis can result in rather severe diseases (e.g., Cushing’s; Addison’s), and attention to these has therefore framed the alleviation of stress. However, given advances in psychoneuroendocrinology and (chrono)pharmacology: efforts to investigate the more subtle neuropsychological aspects of HPA-axis dynamics, such as behavioral regulation, are emerging (ibid., Lightman 2008, Russell & Lightman 2019).
In adrenal insufficiency (cortisol-deficiency), the ‘normal’ function of the HPA axis is bypassed: the **autonomic** regulation of stress by the HPA-axis is overseen instead by the **cognitive** regulation of stress, leading to some interesting (paradoxical) results. The regulation of perceived stress (danger; fight-or-flight) and energy (motivation; ability to cope) does not occur as a silent autonomic process in the background but is under permanent scrutiny due to a patient’s capacity to dose corticosteroids according to what is consciously perceived as necessary and/or appropriate. The bidirectional dynamics between the agent and their embedding are now thus subject to sociocommunicative frameworks in a different way than under “normal” or “natural” conditions. The patient’s **concept of stress**, not just its presentation as the phenomenon of allostatic overload, is now a parameter in the body’s adaptation. The argument explored is that given the sociocultural embedding of concepts, communication with others about the ‘nature’ and definitions of stress now factor into allostatic (self-)regulation, resulting in the preservation of entirely different preferred health-states than the ones neuroendocrinologically given by the body’s own HPA axis. In order to explore explanatory links between the phenomenology of mental allostasis and neuropsychiatric approaches to psychopathology, this chapter seeks to expose aspects of allostatic regulation through a first-person account of corticosteroid self-administration. The author will narrate aspects of their own self-prediction when living with adrenal insufficiency, and this account will be theoretically informed by an allostatic account of active inference and psychoneuroendocrinology. Philosophical introspection will ensue, in every sense of the expression, something which is much needed for updating our understanding of the (predictive) self and (allostatic) health.
Concepts outlive people.^[This statement appears here in order to set the tone, more on this in a few paragraphs. See also: [[Wolfram irreducibility and interconcept space]].] Concepts are the effect of uncertainty-reduction at the social level: we are able to minimize friction—i.e., thermodynamic stress—by assigning (variable) phenomena to categories, hoping these categories remain stable enough for a timeframe of interest. Concepts are therefore dynamic, unstable structures which mediate social scaffolding.^[See: “[[04 Concepts as pre-dictions]]” for the extended take on these thoughts.] Concepts are able to reduce stress, as aforementioned, but they can also create stress: e.g., by triggering or eliciting the *possibility* of stress. From the infamous “don’t think of a polar bear” as a method to bring forth what we could rightly call an ‘intrusive thought’ in someone else’s mind, to ‘the following information will be traumatizing, would you like to hear it?’ These simple examples are meant to set the stage for the claim that, as opposed to the understanding that conceptuality is the domain of abstract and/or representational cognition, **concepts are intimately connected to the physiological stress response**. This bivalence of conceptuality in relation to stress makes it an interesting avenue for exploring *allostasis* as a process with complex feedback mechanisms of adaptive variation, distributed over many scales: from the physiological to the generational to the ideological, and beyond.
The concept of _allostasis_,^[See also: [[Homeostasis]], [[Homeorhesis]] and [[Telepoint]].] introduced by Peter Sterling and Joseph Eyer (1988), represents a paradigm shift in our understanding of stress regulation, moving beyond relatively simple homeostatic models, in order to embrace the complexity of dynamic anticipatory and responsive adaptations. Key to their paradigm-shifting was setting the focus on *systemic* stressors (capitalism, industrialization, psychosocial complexity, urban-dwelling) and how these influence collective and individual health.^[To quote Sterling in a more recent outlet: “Allostasis is “just a fancy name for nothing new ... Whereas homeostasis tends to define “health” as a list of “appropriate” lab values and “disease” as “inappropriate” values, allostasis defines health as the capacity for adaptive variation and disease as a shrinkage or compression of that capacity. Therapeutically, homeostasis emphasizes drugs to clamp down on pesky parameters—two or three drugs to fix blood pressure, other drugs to fix blood glucose, blood lipids, and so on. These drugs tend to reduce the capacity for adaptive variation, whereas allostasis emphasizes the opposite, interventions that enhance the capacity for adaptive variation.” (Sterling 2020, p. xi).] Framed as the adaptive background regulating homeostasis, allostasis can be understood as that which operates on a larger systemic plane of anticipation and appraisal, “triggered by indirect cues beyond the reach of homeostatic control, such as stressors and surprising environmental conditions” (Ororbia and Friston, 2023, pp. 5-6). Allostasis is a concept which tracks a more complex and more contingent projective dimension than that of immediate homeostatic deviations, allostatic processes “finesse homeostasis (or more generally homeorhesis), working to satisfy constraints *before* the need arises; this allows the mortal computer to budget *resources*, as living systems do” (ibid., our emphasis). Importantly for our argument, the “viable ranges” (ibid.) that allostasis is meant to track, are regulated by the given norms of the sociocultural embedding, in the case of human beings. Writing in 1988, Sterling and Eyer (1988, pp. 629-30) explain how and why their research into allostasis took shape, which sets our main interests center-stage: stress as a psychosocial phenomenon, with blood pressure as a variable in stress-exposure (a key domain in the case of corticosteroid effects) and the brain as the regulative link between the two:
 
>No text explains, for example, why in modern society blood pressure rises with age (Eyer, 1975). Nor do they explain why this rise starts at the age when children enter the environment of school ... Texts do not explain why blood pressure is highest and hypertension most prevalent where social disruption is greatest, e.g. among the unemployed and (in the US) among blacks. [In textbooks, these pathologies] are given biological explanations that are unrelated to the social and psychological patterns.
>...
>The only possible link between sociopsychological and physiological phenomena is the brain. Textbooks do not describe this link because the dominant conceptual model in physiology for a century has viewed the body as operating almost independently of the brain.
 
And, we would add: predominantly because our ideological background in W.E.I.R.D.^[*Western, Educated, Industrialized, Rich, and Democratic*, Heinrich et al., 2010.] contexts tends to favor liberal, individualist visions of responsibility. In the context of active inference (AIF), which is our explanatory avenue, the brain-body-cultural niche system is not cast as modularly divided, but as highly interdependent: the complex arrangements of organic systems create constraint-based feedback loops, rather than cause-effect chain reactions.^[Alicia Juarrero shows a variety of compelling examples of this in her 2023 book.] If organic life represents an effort to survive dissipation, these efforts must be constrained by self-preserving principles: something must hold regulative parameters about what *it* is in order to orient itself away from dissipation. This allows us to ask questions such as: “Can the immune response and psychiatric illness both be explained in terms of self-organising systems responding to threatening stimuli in their external environment, whether those stimuli happen to be pathogens, predators, or people?” (Bhat et al., 2021). We seek answers to this question, particularly in the context of people being self-stressing entities, through an account of *semantic adaptivity*, partly from a phenomenology of illness perspective (that of the author).
AIF frames control and planning as _inference_,^[See “[[Active inference]]” for our account.] where _beliefs_ are needed in order to infer (“if X, then Y”). Beliefs, to remind, need not be understood as conscious awareness of a specific representation, but span organic life from the molecular to the social: a bacterium infers itself away towards nutrition by satisfying the “belief” that sugar is desirable, a society holds the ideological background tuning that punishment is an adequate response to offense, and thus guides itself towards the reinforcement of beliefs that jailing practices are acceptable. These and all other observable constraints can be understood as the guiding set-up of a system’s generative model. A generative model, in the case of, e.g., a singular human being, is one that synthesizes experience in tracking sensory data towards what minimizes uncertainty about it in the most self-evidencing way (Parr & Pezzulo, 2021). *Self*-evidencing means that one is estimating incoming sensory data in the way that best explains where/how/what/who/_why_ one is at that moment. In other words, the generative model is that which tracks reality by producing it: no space for naive realism, experimental evidence continues to mount in support of the understanding that no phenomenal access is ever “direct”, it must be generated, and AIF proposes that this generation is best understood as an engaged process of minimizing surprise at every sensorial level, including vast spatiotemporal ranges such as across generations. The brain, therefore, according to Friston, is a “fantastic organ: a generator of fantasies, hypotheses and predictions that are *tested* against sensory evidence.” (2023, p. 257, our emphasis). If therefore, under AIF premises, “normality is context dependent,” (Bhat et al. 2021, p. 26), in thinking about generative models being highly distributed in the case of sociocultural scaffolding, we ought to ask things like: how do concepts, operating across agents, fantastically regulate expected thermodynamic friction, i.e., expected surprise? What kinds of spatiotemporal and/or sociocultural frames are they invested in regulating or preserving or changing? What health and stress parameters in the landscape of psychosocial phenomena which go *beyond* the singular agent? Answers to these questions might change how we approach (psycho)pathology.
For example, though we tend value the type of disconnection characteristic of ‘creative states,’^[More on this in the following section, and in “[[Edging thermodynamic equilibrium]]”.] it is generally accepted that _disconnection_ is an undesirable state: both at the individual level (e.g., disconnection between brain regions leading to schizophrenia, Pettersson-Yeo et al., 2011) and socially (e.g., becoming a recluse, misunderstood, etc.). If disconnection is undesirable, and humans find connection through linguistic-conceptual means, then understanding the mechanics of allostasis as a distributed social phenomenon, tempered and tagged by the everyday notion of “stress”, should be a paramount endeavor. While ‘stress’ remains a notoriously fuzzy concept, for our purposes it can be understood as the ‘tag’ we associate with (the phenomenological manifestation of) resistance to thermodynamic dissipation, i.e., the capacity to deal with highly contingent affairs. It is important to note that this is encompassing both what we *endure* and what we *generate* in response to environmental demands: “... the human stress response has evolved to maintain homeostasis under conditions of _real_ or _perceived_ stress” (Russell & Lightman 2019, own emphasis in italics). Within the complex landscape of stress research, the glucocorticoid hormone _cortisol_ has been the most extensively studied neuroendocrine biomarker of stress, with its dysregulation implicated in a spectrum of psychiatric conditions ranging from PTSD and depression to psychosis and mania, as well as the generalized endocrine disorders of adrenal insufficiency and corticosteroid overproduction. In the case of schizophrenic symptomatologies: elevated cortisol secretion has been linked with increased symptom severity (Walder et al. 2000).
However, more nuanced cortisol fluctuations and their possible links to semantic, social coordination remains an as-of-yet underexplored avenue of research, despite knowledge of its fundamental role in the modulation of physiological, psychological and therefore *social* allostasis. In the context of medically framing something such as mental illness, the trouble lies in determining the spatiotemporal scope of the communicative framework we’re interested in analyzing (how long should ‘abnormality’ be present for, for it to count as _dysfunction_?). A state such as psychosis creates a two-way communication barrier: the psychotic agent fails to connect to certain elements of their psychosocial embedding, and their embedding cannot interpret the agent’s phenomenal reality, even though both may be (internally) consistent. This is where spatiotemporal scales being made *explicit*, through concepts such as allostasis or homeorhesis,^[Homeorhesis, as a concept, was introduced by C.H. Waddington, to describe the tendency of biological systems to return to a particular developmental attractor trajectory rather than a specific state. Unlike homeostasis, which maintains _constant_ conditions, and allostasis, which pertains to adaptive variation, homeorhesis ensures the stability of developmental pathways despite perturbations—metaphorically: in a similar way to how a river’s flow maintains course despite obstacles. This principle helps explain phenomena like canalization in embryonic or growth patterns development, where organisms achieve similar end states despite varying environmental conditions.] is helpful. Allostatically speaking, at the level of the individual, there are no delusional agents: we are all actively-inferring towards our preferred states (see: [[Active inference]]). At the systemic level: this question is one of hegemonic dimensions.^[Even at the level of homeostasis (that is: returning to a setpoint, in facing contingent variation), any understanding of variable interaction and regulation is very complex, very few interacting components are enough to create intractable situations.]
This writing therefore argues for a crucial bridging between psychoneuroendocrinology and philosophy, proposing that the relationship between hormonal and semantic fluctuations is bidirectional and mutually constitutive. A better understanding of this relationship presents implications for both fields: just as conceptual frameworks shape our interpretation and experience of stress, these interpretations in turn influence our physiological responses through neuroendocrine pathways. Therefore, refining our conceptual framing of stress could lead to novel approaches in modulating hormonal responses, while deeper knowledge of endocrine systems might inform our conceptual frameworks. The spatiotemporal tracking of concepts such as ‘health’, through concepts such as ‘stress’, is therefore important to set under analysis: how long is too long, when it comes to stress, and who decides? In a way, this is an argument for furthering psychoanalysis: tuning hormones with concepts, and vice versa, is a phenomenon already leveraged, albeit implicitly, in psychoanalytic practice.
The title of this piece reflects the layered approach analyzed and proposed: the idea of *sick spirit*^[A term borrowed from Gary Gutting writing on Foucault and Hegel, 1994.] emphasizes the psychosocial distribution of the concept of stress, signaling attention to systemic stress pathologies beyond the individual. In their article “Mental Institutions” (2008), Gallagher and Crisafi propose Hegel as a relevant thinker for understanding the implications of the _extended mind_: a recent concept of how the mind can be understood as operating beyond the skull, by Clark and Chalmers (1998). According to Gallagher and Crisafi, Hegel’s treatment of _Geist_, when understood as instantiated through structures such as institutions, is not just revealing of how mind can occur outside the flesh,^[They propose Hegel as moving beyond the “parity principle” proposed by Clark and Chalmers, which is, simply put: if a process, such as memory, can be instantiated through non-brain means, then it can be considered _mindful_).] what the dynamic, dialectical distribution of spirit reveals is that if cognitive extension begins with invention, then in the Hegelian abstraction of Geist in its unfolding, “subsequent acts of cognition are facilitated or enhanced or made possible by particular tools or institutional mechanisms. In every act of cognition that runs through these tools or institutions, the mind is extended” (Gallagher and Crisafi 2008, p. 51): the whole is, in its becoming, much more than just mind externalized. The interest in our exploration is attention to this strange dialectical turn of the screw: (systemic) stress, as a concept, can be internalized, channeled and reorganized through an individual, returning a novel understanding of stress: the sick spirit eats itself and regurgitates: “... the extended mind can come back to bite us; it can place limitations on our thinking, as easily as it can enable great and wonderfully extended cognitive performances” (ibid.). “[H]ow is ... language, which is no longer that of anyone, which is being’s universal self-consciousness, to be distinguished from human, all-too-human language? In other words how does the passage from Phenomenology to absolute Knowledge work?” This, Gutting tells us Hyppolite tells us: “is the Hegelian question par excellence” (1994, p. 23).
The opening citation by Duerr^[Thanks to Gabriel Catren for this reference. Also, writing about this citation, Haller writes: “We are in no position to assert that for alien thought to be made intelligible, it should be required to synchronize with our own concepts and conventions. As Peter Braun, the literary analyst of ethnology, describes this position of Duerr’s, “There is no authority…neither experience itself, nor a universal ‘logical grammar’ or comparable axiomatics of rationality, with the help of which the question could be decided and therefore different world-views could be compared and tested for their truth content” (Haller 2012, 275).”] gets at this disorientation: there is no measure for this. However, and Foucault reminds us of this too (1982, p. 779), formalizations are needed in order to better understand how they do not work. This is how cognition innovates by active inference, how all models are wrong but useful: “if not that, maybe this....”. The permanent metastability between sanity and madness in the dialectics of this trial and error is a strange self-correcting enterprise (Sellars): hypothesis-testing—in science or in actively inferential cognitive processes—is premised on the assumption that some beliefs need to be assumed or taken for granted in order to test others, that this is a distributed process in linguistic agents, and that all beliefs are subject to revision. This is “self-correcting” in the most erroneous and recursive way,^[As Hui explains: “Recursivity is not mere mechanical repetition; it is characterized by the looping movement of returning to itself in order to determine itself, while every movement is open to contingency, which in turn determines its singularity.”] inspired by the framing of our speculation through active inference (AIF): thermodynamic frictions which lead to allostatic adaptations are all about error-prediction: in order to know where I will be next, I certainly know, at the very least, that I will not be where I am now. Framing things in additional psychoanalytic light, as Carthart-Harris et al. (2014) explain: Freud argued that “dreaming and psychosis typify a primitive style of thinking that is dominant in human infancy” (p. 7), primitive because it lacks recursive reflection, or metacognition: the ability to reflect our own thoughts and behavior (Fleming et al., 2012). Extending the mind into its Geistig consequence, we can frame a kind of metametacognition occurring at the psychosocial level: the distributed concepts which temper our sanity are just as susceptible to primitive dreams and delusions as are the minds of those we dismiss as incomplete (children, the ‘mad,’ and others).
In what follows we will open with an exposition of the idea of “scales of sanity”: how sick is too sick? Then we will analyze the concepts we are treating, in order to conclude with the idea that stress, health, and other general “public domain” concepts, can be better analyzed in terms of the spatiotemporal scales they can be understood as tracking.
 
### Scales of sanity: how sick is too sick?
 
>The idea of white supremacy rests simply on the fact that white men are the creators of civilization (the present civilization, which is the only one that matters; all previous civilizations are simply contributions to our own) and are therefore civilization’s guardians and defenders. Thus it was impossible for Americans to accept the black man as one of themselves, for to do so was to jeopardize their status as white men. But not so to accept him was to deny his human reality, his human weight and complexity, and the strain of denying the overwhelmingly undeniable forced Americans into **rationalizations so fantastic that they approached the pathological**.
>
>J. Baldwin, “Stranger in the Village.” _Notes of a native son_, 1955. Own emphasis in bold.
 
The link between phenomena we tend to designate as ‘madness’ and ‘creativity’ may be obvious, and the line separating the two: incredibly difficult to draw. An approach we _can_ take in order to link both phenomena is the psychosocial intelligibility of their respective inferential chain processes. On the one hand: creative acts (a new mathematical proof, a painting, a good joke) couple with the systems that enable them by—if ever so slightly—challenging established norms (understandings, procedures, styles), and possibly inventing-revealing novel ones. This is sometimes accepted synchronically: a community of agents interprets and understands the creative act as it takes place, but sometimes diachronically: it takes time for an idea to be accepted and understood. Note that we speak of agents and/or their creative ideas as having been “ahead of their time” when the latter happens, implying the spatiotemporal scaling—chunking and parsing—analysis we are interested in underlining here.
Entertaining this link reveals that there can be said to exist some sort of societal consensus that we are talking about the relativities of cognitive _speed_, i.e., how predictively effective an idea is, and how this implies its psychosocial embedding as acceptance and use. From an AIF perspective, ‘delusions’ (i.e., moments of madness) can be framed as ‘errors’ of precision estimation. However: which side of the mirror are we looking at or from? Friston notes that “when one thinks about psychiatric and neurological disorders, most can be framed as false inference” (2023, p. 257). But this seems restricted to the domain of the individual: a person in a delusional state. What about larger dimensions? The background, semantic normalcy determines the acceptance rate of the errors, and as we noted: this seems to be about the tolerance rates of relative speeds; chunks parsed in a particular manner. In fact, it seems we sustain grand-scale inferential havoc if we think of many aspects of our unfolding histories.
Even at the level of the individual, thinking about phenomena such as paradoxes and illusions, which are the hallmark of phenomenal instability:^[For an extended exposition of paradoxes and AIF see: [[04 Concepts as pre-dictions]].] if these are the result of representations which come into ‘contradiction’ with each other when they can be perceived at once (e.g., convex-concave confusions), an interesting thing to point out here is that the perceiver is changing _who they are_, spatiotemporally tuning their relative condition, with regard to what they are experiencing. This is highly revealing of how the imbuing of phenomena with meaning—actively inferring—is a pretty automatic effect,^[Which can be understood as free-energy minimization in the interest of complexity-reduction for better self-actualization.] and therefore: *all* affordances we experience are an effect of the a body’s encounter with inevitable, irreversible dissipation, as it tries to keep relatively stable the illusion that it remains the same entity through time (this is what leads to perceptual disambiguation). What is even more interesting in the case of the type of adaptive creature we are, is that with a little effort we can ‘teach’ ourselves to sustain ambiguity (e.g., in optical illusions, or in poetry), which allows for an ever larger structure of predictive meaning (if more is possible: more is possible). The rabbit-duck or duck-rabbit is not a rabbit or a duck anymore, it’s the rabbitduckrabbitduck.... Not all illusions permit this, but many do, and it’s remarkable to witness this taking phenomenal effect. As we enlarge our cultural niche (or Geist enlarges us), creatively enrich our contexts, what we seem to be doing is allowing more and more space for ambiguity. Ambiguity elicits interpretation, which elicits cognitive effort, which leads to innovation: “if not this, then what?” The perplexed mind, creative or mad, is capable of witnessing layered ambiguities of high complexity. The expansion of Geist, when understanding its dialectics as the permanent reorganization of contradicting affairs, seems to reveal an increasing tolerance for, and engagement, with ambiguity. Navigating complex ambiguities demands sophisticated cognitive engagement (which, if creative, should be able to be highly associative: connective),^[If, as most AI experts seem to agree upon (e.g., Chollet 2019), intelligence is about cross-domain; all-terrain capabilities, then associativity should be a hallmark here, too.] this creates a productive tension between what is known and what remains uncertain. In terms of perception being actively inferential: holding multiple possible interpretations simultaneously reveals pathways to challenge existing models and frameworks, forcing the transcendence of established approaches.
Acts we relate to madness (think: the mad hatter, psychotic breaks, norm-disrupting behavior, “momentary lapses of reason”) are as creative as societally-accepted bursts of “pure reason,” but the **interpretability** and **speed** of their inferential procedures seems different than that of creative moves. They follow a similar—again, if not the same—logic, but appear to be undergirded by different constraint regimes, mostly in terms of spatiotemporal spans. That is: how much effort, as time, or which location, is needed to interpret and enable their appearance on the communicative frameworks that (could) sustain them. E.g., the psychotic agent often connects patterns which others are not able to follow (these are interpreted by the community as paranoia, delusion, etc.). Do note that the psychotic agent conceives of themselves as synchronic and coherent (most accounts of psychosis are revealing of this effect, see examples in Kusters 2020). However, from the perspective of a psychosocial community, their generative model is overestimating certain details of an otherwise “agreed upon” consensus reality. The errors are seen to be stemming from within the agent, whereas in creative (e.g., artistic or industrious) acts the errors appear as novel, explorative dimensions on an already-accepted social domain.
The speculation explored in this brief exposition, in order to frame how we track spatiotemporal embeddings through regulative concepts, is this: it seems so-called madness is faster and smaller-scale than creativity, sometimes residing within a single agent (in terms of communicative frameworks: so fast that it sometimes leaves everyone behind, again: see Kusters 2020). Madness can be fleeting or sustained, but is often fleeting and therefore dismissed. Creativity can be just as fast, fleeting and singularized, but couples with its sociohistoricomaterial constraints (i.e., with Geist) with longer-lasting effects, whether synchronically or diachronically. It’s only the ideological, normative, etc., aspects of these constraints which enable creative acts to last, which leads them to be differentiated from “mad” acts. However, both types of acts, creative and mad, can be understood as the same process, it’s just that their effects are interpreted and occur at different spatiotemporal scales, against a regulative background.
Examples linking these phenomena abound: all _avant-garde_ artistry is an effect of dislodging from the old. This means: away with (local) tradition, and onwards with something which challenges its norms. Rules and roles change all the time. In the “Western” context, Duchamp is the poster child example for 20th century madness-creativity, or _Le Salon de Refusés_, slightly earlier.^[Although the glorification of Duchamp’s persona for this gesture has recently come under scrutiny, given discoveries about the idea having originated in the creative madness of Elsa von Freytag-Loringhoven.] Antonin Artaud, James Joyce, many others. Also, clearly: jesters, witches, troubadours, etc., through the ages. In artistically-explored mathematics the prime example of the madness and psychedelia of transformations is Lewis Caroll’s _Alice_. “Eureka” moments designate outbursts of madness which is psychosocially enabled as creativity. *Eppur si muove*: the opposite. Or, in our time: the drinking of milk which is not from one’s own mother: pretty damn crazy. All museums today house a vast diversity of illegibly mad phenomena, made legible by the protecting and isolating containment (Markov) blanket that is the museum. Its border keeps the madness at bay, and allows for a psychosocially accepted meta-perspective on these phenomena: it allows for them to be seen-_as_ and not just _seen_. The tortured artist/creator/thinker: the very image of creativity; the witch’s brew.
Hopefully this exposition clarifies some of the paradoxes at play, we now move onto how the psychosocial distribution of concepts begins at the organic distribution of cognition over an entire physiology.
 
### Paradoxes of stress-regulation through corticosteroids
 
>... dysfunction of an endocrine gland does not create disease or illness, it creates a changed human being.
>
>Karl König, _Meditations on the Endocrine Glands_, 1952.^[I am no anthroposoph, but this had to include this citation in this context.]
 
>The repeated discharge of corticosterone provokes either functional disturbances such as vascular spasm and hypertension or morphological lesions such as stomach ulcer. Hence in the populations of English villages subjected to air raids in the last war, a notable increase in cases of gastric ulcer was observed.
>
>Canguilhem, 1950, p. 30.
 
In the particular condition of adrenal insufficiency with which the author lives,^[I manage stress by coping, like everyone else. However, like all adrenal patients, as will be explained, I have a different access to my stress response: I cognitively modulate it. Instead of cortisol being released by the experience of stress (arousal, surprise, phenomenal frictions of any kind), I consume hydrocortisone tablets (hydrocortisone is synthesized by the liver and becomes bioidentical to cortisol) in order to mimic the “natural”, autonomic stress response. This means, for example, that I take hydrocortisone at 3:30am, every night, in order to mimic the waking cortisol curve. I then proceed to take it at the rest of the “natural” peak hours, in order to function as physiologically normally as possible. However, I cannot possibly know, except in very rough averages, how much cortisol is needed for _additional_ stress. Think of an exam, an emotional situation, work stress, political stress, sport stress, etc. I resort to averages which help me guide myself through the concept of stress. The standard approach for very stressful situations (a car crash, sustaining intense injury, etc.) is 100mg of hydrocortisone. This is about 5x or 6x the cortisol amount a body produces in one day. But things can be subject to change: last time around I broke my ankle I actually needed 200mg to regain consciousness (I was neither here nor there, I was sort of catatonic). It remains a constant mystery and a full-time side job to figure out how much and when to dose.] the normal functioning of the HPA axis is bypassed: the autonomic regulation of stress is overseen instead by the cognitive regulation of stress, leading to some interesting (paradoxical) findings. The regulation of stress and perceived energy—motivation, ability to cope, etc.—does not occur as a silent process in the background but is under permanent scrutiny due to the author’s capacity to dose hydrocortisone according to what is perceived necessary and/or appropriate. This confuses allostatic adaptation, since, as Clark suggests: when “bodily ... self-monitoring ... is somehow malfunctioning: [one] will under- or overestimate one’s body’s present and future needs” (Clark 2023, pp. 156-9). The vital dynamics of cortisol emerge from feedforward and back-regulation between the pituitary gland and the adrenal cortex (Russell et al. 2023). As the end-product of the hypothalamic-pituitary-adrenal (HPA) axis, cortisol is produced in *anticipation* as well as in *response* to stress. This dialectic sets the focus on how the “old distinction between the “psychiatric” and the bodily/physiological/neurological needs to be abandoned.” (Clark 2023, p. 156). Clark, Barrett, and others (Arnaldo et al. 2022) also suggest that if conditions such as chronic depression involves “abnormalities not simply of “mood” but also of sleeping-waking cycles, and of metabolic and immunological response ... [then they] may be a “central problem with inefficient energy regulation. Mistakes in energy budgeting would normally be corrected by prediction error signaling” (Clark 2023, p. 158). What is energy-efficiency if not the adequate (i.e., comfortable; leading to future comfort states) processing of contingency? What is comfort if not the current status quo? Sociocultural embeddings of stress concepts directly influence allostatic regulation, leading to the preservation of different preferred health-states (i.e., to remind: Geist is more than just the extended mind) than those naturally mediated by the hypothalamic-pituitary-adrenal (HPA) axis, especially in the case of the author (but the same applies to all of us: I am just here to reveal the mechanics). If normalcy is relative (as proposed by AIF, and much earlier by Canguilhem), we ought to tune concepts in order to fine-tune health and measures of survival.
Again, in adrenal insufficiency, where the normal functioning of the hypothalamic-pituitary-adrenal (HPA) axis is bypassed, the autonomic regulation of stress is replaced by cognitive regulation: the autonomic stress-response process is replaced by conscious supplementation through hydrocortisone^[Or in some cases another steroid: prednisone, dexamethasone.] tablets, which are metabolized by the liver to become bioidentical to cortisol, leading to a rather specific experimental condition, where one must explicitly engage with allostatic needs being met, translating perceived stress into hormonal requirements. The most immediate paradox to point out here is the _dosing paradox_: the very stress of managing stress requires additional hormonal support, creating a potential positive feedback loop. This increases the difficulty in distinguishing between perceived and lived physiological stress: when symptoms are ambiguous—e.g., nausea, dizziness, or lack of energy—one must engage in a complex interpretive process to determine whether these symptoms indicate insufficient cortisol levels from past stress, or signal an ongoing (or perceived future) stressor requiring increased dosing. The concept of self-identity becomes similarly problematized when one can actively modulate one’s hormonal state: the familiar phrase “not feeling like yourself” takes on new meaning when one has direct access to the hormonal modulators of mood and motivation. This puts the author in permanent, profound contemplation of existential questions about the nature of authentic experience and the role of neuroendocrine function in personal identity. The paradox deepens even more when considering that corticosteroids can have euphoric psychological effects, making it difficult to distinguish between appropriate supplementation and over-replacement.
As we’ve been arguing: the dialogically-distributed nature of stress as a concept becomes particularly evident in this acute context. Hormone-administering patients need to constantly reference others’ experiences with stress to calibrate their own responses, yet this very process becomes part of the stress regulation system itself, and therefore: the sociocultural embedding of stress management cannot be divorced from its physiological implementation. This may suggest, depending on which avenue we take, that the concept of stress itself can function as a marker of allostatic adaptation in human systems: people’s working definitions and narratives of stress, paired with cortisol measures, could in the future provide insights into how society as a whole adapts to (self-)generated stress conditions. Following our quest for a more refined metapsychoanalysis: this understanding breaks distinctions between physiological and psychological conditioning. If the biochemical etiology of allostatic overload results from *perceived* interoceptive and/or *environmental* stress (McEwen 2005, p. 316), and if endocrine mediators like cortisol and adrenaline both respond to and contribute to allostatic overload (ibid.), then the process of societal self-stressing represents a dialectic where the sociocultural embedding of concepts means that communication with others about the nature and definitions of stress now factors directly into allostatic regulation, resulting in the preservation of entirely different preferred health-states than those neuroendocrinologically given by the body’s own HPA axis: adaptation to self-produced stress changes the nature of what is considered stressful.^[This is, in broad brush strokes, how we become insensitive or hypersensitive to specific phenomena.] If psychopathology is ‘false inference’ (Friston 2023, p. 257), and the mechanics of the generative model can be observed in a logic, which from the outside, is seen as aberrant belief-effectuations as they couple to a larger communicative system (e.g., the social one): what is to be said of the mechanics of _systemic_ psychopathology, which we cannot see from the outside other than through our concepts?
Despite decades of research on the psychiatric symptoms associated with corticosteroids (Sirois 2003, p. 28), and clear evidence that corticosteroid-excess and deficiency induce neuropsychological changes even in short-duration therapy (Russell et al. 2023), an account of how they affect the prediction of an allostatic trajectory beyond mere homeostatic health has not been studied.^[This has mostly been since dysfunction of the HPA-axis can result in rather severe diseases (e.g., Cushing’s; Addison’s).] Recent advances in pharmacology have enabled more nuanced investigations into the subtle neuropsychological aspects of HPA-axis dynamics, including behavioral regulation (Lightman 2008; Russell & Lightman 2019). These investigations can and should be combined with the type of semantic research proposed here, to underline how the relationship between hormonal regulation and conceptual frameworks of stress is bidirectional and deeply embedded in social and cultural contexts. This could imply a fundamental revision in how we understand stress regulation: the boundary between “natural” and “modified” stress responses becomes increasingly blurred, as does the distinction between physiological and psychological stress. A new image of stress emerges, as an evolutionary phenomenon of ever-more-complex entropy-management that operates at multiple levels—physiological, psychological, and social—with each level influencing and being influenced by the others.
### Phenomenology of own case
 
>Oxytocin, serotonin, codeine, cortisone, the estrogens, omeprazole, testosterone, and so on, correspond to the group of molecules currently available for the manufacturing of subjectivity and its affects.
>
>P. B. Preciado, _Testo Junkie_, (2008) 2013, p. 118.
 
The paradoxes of self-dosing stress hormones (which also includes aldosterone in my case) emerge in the grey zones, not at the extremes. Most patients go by “strictly physical” symptoms in order to adjust their doses, this is also what is advised by most knowledgeable endocrinologists.^[A lot of not really caring ones simply imagine that 20mg is all anyone needs, and don’t even inform patients about the highly dynamic chronobiological aspects of it all.] So, let’s say one day I feel slightly nauseous, dizzy and lacking energy (as mentioned above: these are all symptoms of low cortisol). There are a few interesting observations I can make as a patient here: 1) something _specific_ happened which drained my cortisol (stress I didn’t realize was actually pretty stressful, usually a day or few days before), 2) something less specific and more lasting is happening and I cannot yet know what it is (prolonged and sometimes increasing stress, caused by, e.g., an underlying issue such as an infection, or interpersonal emotional stress, increased physical exertion, etc.). In either case I need to temporarily increase hydrocortisone to resolve my symptoms. Usually, with an increase, the symptoms are resolved. However: in the case of prolonged stress (e.g., work pressure), when should one address the pressure, rather than the physiological coping? This can only be done by assessing the concept of stress itself. What would others find stressful? When would they retreat and take a few days off and stop working? How to assess all this in a sociocultural paradigm so work-driven and allostatically dynamic? Paradoxical, to say the least.
In other cases, relatedly, the sense of self is another grey zone which I modulate with hydrocortisone. “Not feeling like yourself” is a familiar phrase most of us will recognize as a returning experience. Sometimes things are “off”, you’d like to be in a different mindspace, mood, motivational inclination, but you’re not. And you cannot pull yourself by the hormonal bootstraps, but in my case I sometimes can. But the paradoxical thing about this is: what is my “good” sense of self? When do I achieve this? I most often like to feel helpful, ready, capable, activated, friendly, productive. When I don’t feel these things I don’t feel like myself.^[This is painting an overly simplistic picture for the sake of argumentation.] But, the catch-22 is that corticosteroids can have euphoric psychological effects, so these characteristics can be easily induced by increasing their consumption. I have certainly felt this at times, which made me temper my increases. At the same time, the opposite effects can also be true: too much hydrocortisone can induce a sense of overwhelming desperation, which is extremely unpleasant. But, punishingly so, too little cortisol can induce both these extremes, too.^[Something rather funny in retrospect, but absolutely not funny in the moment itself: as I finished this chapter, January of 2025, I was undergoing severe symptoms of hypercortisolism, which made me _extremely_ paranoid and unwell. Lacking the knowledge that what I needed was extra fludrocortisone, which replaces aldosterone, I simply kept increasing my hydrocortisone because of the stress I was experiencing, which only made things worse. Part of the medical explanation for this is that while the “secretion of the stress hormones adrenalin and cortisol in response to an acutely threatening event promotes and improves memory ... however, when the stress is repeated over many weeks [in my case: too much hydrocortisone keeps being administered repeatedly], some neurons atrophy and memory is impaired, whereas other neurons grow and fear is enhanced.” (McEwen 2005, p. 317). I live with the additional stress, all the time, that I am atrophying my capacity for thought. This is additional stress. This is the permanent catch-22 of the concept of stress. See also Sapolsky 1994, pp. 109-113 on stress and memory.] The confusion, then, when symptoms are variable, is large. Do I feel bad because I took too much? Or do I actually need more to address how I feel? Hypothesis-testing is all I have at these moments, which results in a scrutinizing conceptual assessment of the concept of stress itself. Moreover, as a lot of patients also lament: stressing about stressing and dosing signifies extra stress which may in turn mean more dosing.^[According to a 2010 study on the stress of patients with Addison’s disease (cortisol underproduction): “patients with Addison's disease have not only increased levels of anxiety and fear, and over-reaction to stimuli, but decreased performance efficiency and need for social contact as well.” (Warmuz-Stangierska, Baszko-Błaszyk, Sowiński 2010, p. 90). This is indicative of what has just been outlined.] This is a strange positive feedback loop where a lot of things are prone to go awry.
Again, Clark argues that if predictive processing is on the right track “depression is never simply a disorder of mood. It is a disorder of the whole body-brain-environment system, affecting the way the brain forms and runs bedrock bodily energy budgets, and the way it responds to new positive and negative information.” (Clark 2023, pp. 162-3). We could not agree more regarding our dialectical process philosophy of the concept and phenomenon of stress.^[In depression, Ratcliffe explains that the sense of the possible, anticipatory, predictive capacities deteriorate (2015) As explained by Kahn (2024): “The perceptual awareness within which one anticipates possibilities is diminished, and as a result, things don’t matter anymore. Experiences of depression involve a changed way of being in the world; a profound existential shift with far-reaching consequences occurs to a background layer of experience. Because access to various kinds of possibilities is constrained, meaningful engagement with the environment is forestalled.” (Khan 2024, p. 5). These insights are very revealing in the context of cortisol, which is strongly coupled with the experience of time and the ability to cope (i.e., adequately manage energy-budgeting) towards self-evidencing.] The experience of patients with adrenal insufficiency, who _consciously_ manage their stress response, provide insights into the interactions between _hormone-concept-systems_. The interpretation and management of physiological states depends heavily on socially-mediated concepts and environmentally-distributed information processing. As Zwart (2008) shows, _heteroglossian_, polysemous understandings of phenomena are what lead to the social, philosophical and scientific refinement of concepts through the dialectics of historical unfolding (pp. 77-97). While stress, as the experience of coping with thermodynamic duress, may be localized within what we can call an _agent_—much as the mechanics of combustion and locomotion are localized within a vehicle—the higher-order phenomena that emerge from these basic processes cannot be easily pointed at or localized. Just as the concept of “travel” cannot be located within the mechanical operations of a vehicle, allostasis cannot be reduced to or located within the neurophysiological operations of an organism: the idea of a “ghost”, might be effectuated as fear _within_ an agent, but it effectively came from “the outside”, this can be understood as throwing in/out distinctions out of the window, revealing a new ‘topology’ for understanding psychosocial adaptation. This has implications for understanding how stress regulation operates across multiple scales of organization, from the molecular to the social, and suggests that any complete account of stress must consider its emergence from and embedding within complex networks not only of biological and psychological, but crucially of evolving social processes.
As Hannibal and Bishop argue in the context of pain management therapy, while stress may be unavoidable: being embedded communicative beings with the capacity to regulate each other, we also have the capability to modify what we perceive as stressful and how we respond to it:
 
>Exaggerated psychological responses (eg, catastrophizing) following maladaptive cognitive appraisals of potential stressors as threatening may exacerbate cortisol secretion and facilitate the consolidation of fear-based memories of pain or non–pain-related stressors; however, coping, cognitive reappraisal, or confrontation of stressors may minimize cortisol secretion and prevent chronic, recurrent pain ... and improve quality of life.
>
>Hannibal and Bishop 2014, p. 1816.
 
These observations also link directly to active inference research on optimism bias, where the amplification of optimistic outlooks seems to prove that we are also capable of modulating optimism itself (Fisher et al., 2024). As famously presented by Sapolsky (1994): we are creatures who, sometimes out of mere boredom (the phenomenon where need for challenging contingency presents: we want to maximize information-gain), _create our own stress_ (e.g., pp. 136-140).
### Conclusion: collective allostatic inference
 
>Deleuze takes up what the Stoics call a quasi cause, which is not a real cause in the sense that the relation between cause and effect is not necessary. In other words, there is no necessity between getting sick and becoming a good philosopher, but getting sick can be an event qua quasi cause that transforms the person into a good philosopher (for example, Nietzsche): Being sick implies a double causality: one is a corporeal cause, the other an incorporeal quasi cause: “The Stoics saw clearly that the event is subject to a double causality, referring on one hand to mixtures of bodies which are its cause and, on the other, to other events which are its quasi-cause.” See Gilles Deleuze, _Logic of Sense_, trans. Mark Lester (New York: Columbia University Press, 1990), 94.
>
>In: *Recursivity and Contingency*, Hui 2019, p. 40.
 
In our account, the Geistig, social dimension of stress regulation as allostasis manifests in how hormonal responses propagate through collective semantic interactions. There is no denying that we produce hormonal flushes in each other through our sociolinguistic interactions: this seems to be one of the most obvious ways to phrase the way in which hormones tune concepts, and vice versa. Following an actively-inferential account, this is a phenomenon that can be understood as a collective form of uncertainty minimization in group dynamics: stress responses emerge not just from individual physiological processes, but from the continuous mutual prediction and adjustment occurring between socially embedded agents. Following the impetus set behind the concept of allostasis by Eyer and Sterling: instead of focusing on predetermined, prestatable behavioral rules, we need to think about what parameters an organism tunes and is tuned by in relationship to its environment to explain its adaptive behavior. If social behavior is a process of self-evidencing through extended means, where systems continuously refine their generative models while actively changing their states to minimize prediction error, this suggests that in the complex interplay between individual physiology and collective dynamics (where individual hormone regulation becomes part of a broader social process of uncertainty minimization): the fundamentally distributed nature of what we typically consider, medicate and treat as individually pathologized responses, ought to be radically challenged. Instead of hearing “here is your prescription for blood pressure medication” at the doctor’s office, we might like to hear: “have you thought of the systemic oppressors that have your numbers up?”
Again, the parameters which determine a state of (mental) ‘health’ are debatable, if not altogether impossible to determine. Dynamic feedback cycles of self-regulation ground the vital and functional conditions of organisms, and it is precisely because organisms display repeating tendencies towards specific sates, that we can speak of _homeostasis_. However, the idea of homeostatic function alone cannot account for the adaptive responses to cycles of stress, especially when they are self-generated. _Allostasis_ represents the capacity to “maintain stability through change”, which in the context of predictive processing we can certainly frame the ways in which persistent organisms alter their models of expected stability to adapt to changing needs (Clark 2023, p. 137). Allostasis as resource-management, as permanent systemic stress-analysis, ties in a whole range of currently topical concepts: resilience, resistance, robustness, plasticity, buffering, to name a few.^[>“Lyotard was right when he said that modernity should be understood not as an epoch but rather as the capacity to tolerate contingency, which is called resilience today.” Hui, 2019 p. 25.] All of which are strongly tied to images of _adaptive_ intelligence.
The philosophical conjecture could be made that stress, the concept, simply signals what we currently understand as energy-regulation against existing frictions, and the accepted degrees of regulation are subject to social organs which cannot be decoupled from internal physiologies. As mentioned, remaining within viable energetic bounds through glucocorticoid feedback cycles is key to allostatic adaptation (Sapolsky 1994, 2009, McEwen 2005), and allostatic ‘disorders’ pertaining to “mistaken forms of bodily prediction involving energy regulation” (Clark 2023, p. 156) pervade all aspects of modern life. “How did ... maladaptive responses [to stress, resulting in corticosteroid overload] evolve? The most likely explanation is that the body simply has not evolved the tendency _not_ to secrete glucocorticoids during a neurological crisis” (Sapolsky 1994, p. 117). We proposed to look at the “mistakes” in energy regulation resulting in varying degrees of prediction-projection through the lens of cortisol: taking from personal experiences with the use of corticosteroids to regulate adrenal insufficiency, the argument made that sensing interoceptive energy-availability, which is primarily regulated through cortisol, leads to variations in ability to predict through and for conceptual coupling to larger systems, which results in ‘paradoxical’ feedback loops which involve the sociosemantic evolution of the concept of stress itself.
Recent advances in biological physics provide compelling support for a hermeneutic understanding of organism-environment relationships. As Bhat et al. (2021) argue, living systems engage in constant energetic dialogue with their environments, a process that is inherently ‘imperfect’, enactive and contextual. Traditional notions of stability in biological systems are out of the window, the implications of this extend beyond mere philosophical speculation: the relationship between neuropsychiatric suffering and semantic modulation reveals itself, therefore, as constraint-based and recursive rather than straightforwardly causal. All mental states reflect their extended nature as they do not exist in simple linear relationships but in complex recursive networks. This understanding helps explain why semantic interventions such as psychoanalysis can have profound effects on neuropsychiatric conditions: the evolving nature of these relationships frames meaning-making processes beyond the epiphenomenal. Contemporary approaches to psychopathology reveal a stark contrast between individualized medicine’s focus on personal dysfunction and the broader phenomenon of collectively sustained delusions. As Jylkkä (2024) notes, societies often maintain comforting but potentially dangerous collective beliefs that serve stabilizing functions, which brings us to fundamental questions about the nature of mental health: can individual psychological wellness be meaningfully separated from the broader ecology of social beliefs and practices which rest on medical images of homeostatic stability? As Eyer and Sterling note: homeostatic models do not scale well, and stressors, in general induce the mobilization of energy. This leads us to question the home(ostatic) economics of our current (mental) health landscape. What kinds of spatiotemporal chunks are we prioritizing?
The relationship between stress and entropy measures (of the brain, and extended sociocognitive systems) can therefore offer a window into systemic adaptation. Research by Carhart-Harris and others has demonstrated complex interactions between serotonergic systems and stress hormones like cortisol, suggesting that mental states can be read in terms of their entropy characteristics. As we explore elsewhere:^[See: [[Edging thermodynamic equilibrium]].] much remains to be studied here. The spatiotemporal dynamics of concepts pose analytical challenges for naturalizing metaphysics, such as the one that may be possible via AIF. The various presentations and conceptualizations of stress will always remain perspectival, and the models of science will inevitably embed themselves in the everyday expressions of the stressors of life. If concepts emerge from and are sustained by particular configurations of materiosocial systems, should we understand them as having ‘natural’ lifespans or expiration dates?^[This question inherits much from the not-yet-expired ideas of Gregory Bateson, who introduces “Steps towards an Ecology of Mind” (1971) with: “How do ideas interact? Is there some sort of natural selection which determines the survival of some ideas and the extinction or death of others? What sort of economics limits the multiplicity of ideas in a given region of mind? What are the necessary conditions for stability (or survival) of such a system or subsystem?” (p. 1).] All agents have, after all, different acceptance rates of _errors_. Can we move to different conceptualizations of stress if we understand all agents as tracking a different *types* of complexity? If complexity—which leads to novelty, creation, exploration—represents higher entropy, and therefore more ‘stress’: what are the limits of physiological (and psychosocial) tolerance here? These questions become pressing in an age when we have to consider how rapidly the material and social conditions that give rise to concepts seem to change. A truly naturalistic approach to concepts might need to incorporate explicit, formal consideration of their spatiotemporal dynamics and the conditions of their persistence and transformation (we treat aspects of this in [[11 Post-Control Script-Societies]]).
<div class="page-break" style="page-break-before: always;"></div>
### Footnotes