---
title: "Experience without a Chair"
language: en
date: 2026-09-11
canonical: "https://andreasvieth.de/en/materialien/animal-ethics/experience-without-a-chair/"
format: markdown
generator: "TYPO3 LLMs.txt Extension"
---

# Experience without a Chair

Export the page as structured markup – Optimised for machine processing and re-use. (Use the browser back button to return to the website view.)

 [ Markup lesen ](https://andreasvieth.de/en/materialien/animal-ethics/experience-without-a-chair/.md.md)

## Experience Without Voice

##### **1. Introduction: What happens on this page**

This page does not address the question of which beings count morally. It addresses the question of at which points a being that already counts can vanish from moral attention – without anyone disputing that it counts.

The starting point is an asymmetry. Animals are not expressionless, but they are not discursively self-representing: they can flee, refuse, avoid and cooperate, but they cannot answer an interpretation with a counter-report. From this follows the methodological guiding thesis of the entire page: **Where propositional self-representation is absent, representation must not become freer but must become more demanding.** It is not moral significance that decreases, but the burden of scrutiny that increases.

The second methodological decision concerns the concept of de-addressing itself. It does not denote every act of selection. Every perception, every experiment, every category, every statistic and every procedural order selects; thought without abstraction would not exist. The criterion lies one level deeper: de-addressing arises where a description **blocks the return path of what it screens out**. This limits the charge while also making it testable – it is directed not against abstraction but against its becoming self-sufficient.

The third decision concerns form. The page does not argue linearly from a principle to its consequences, but **topologically**: it maps places where the return path can be severed. The foundational section is followed by eight such layers – body and expression, the situation of data collection, conceptual stock, discourse pragmatics, the chain of translation, practice and arrangement, institution, theoretical form – followed by an extension to species and ecosystems, and finally a meta-norm running across all layers. The layers are neither a causal chain nor a hierarchy. They are places, and each place can be intact or blocked on its own: an animal can be perceived with phenomenal precision, surveyed with methodological openness, described adequately in conceptual terms – and yet be represented institutionally in a way that remains entirely without consequence.

Three distinctions carry the construction. First, **transition control**: experience, moral relevance and practical decisiveness are not the same thing. That an animal feels pain states how it fares; that this pain constitutes a moral reason is already a normative claim; that it is decisive in a given situation requires further bridging principles. The arrows between these stages are not automatisms but points of examination. Second, the **typology of blockage**: it can be *privative* (a concept, a procedure, a channel of revision is missing), *preclusive* (something already present occupies the space and no longer permits anything else), or *constitutive* (the instrument of knowledge itself generates the blind spot). Three different repairs correspond to these: supplementation, unblocking, methodological critique. Third, the **ridge structure**: every counter-operation has its own characteristic failure form. Whoever avoids anthropocentric silencing readily ends up in paternalistic advocacy; whoever criticises concepts, in conceptual decree; whoever demands visibility, in panopticism. The page therefore nowhere recommends a compromise between two errors, but rather a controlled passage between them.

On the handling of sources: Wittgenstein, Buytendijk and Plessner, Despret, Latour, Fricker, Langton, Donaldson and Kymlicka, Diamond, Crary, Luhmann and Peirce are invoked not as authorities supporting a position, but each for a specific **diagnostic function** – and where their concepts fail to carry, the limit is marked. Fricker's hermeneutical injustice is, for animals, an analogy, not a subsumption; Langton's *blocking* is the counter-operation and precisely not the blockage; Luhmann provides a diagnosis of societal selectivity, but no environmental ethics. These demarcations belong to the matter at hand, because a topology that fails to control its own borrowings commits exactly the error it describes.

The page therefore does not end with a principle, but with a norm governing principles: **Addressing does not mean screening nothing out. Addressing means keeping open the return path for what has been screened out.**

##### **2. Cursory pass-through: the traps and their avoidance**

**Body and expression.** The trap lies not in measurement but in the transition from methodological reduction to ontological reductionism: an animal that does something becomes a body to which something happens – the "dead melody." Indicator: what has been screened out may in principle no longer return. The counter-operation is not the renunciation of data, but their re-linking to the behavioural gestalt from which their variables were abstracted. The failure form of this repair is unexamined anthropomorphism, which replaces the emptied description with a humanised one.

**The situation of data collection.** A setting de-addresses when it permits the animal only to reproduce the presuppositions of the experiment – docility instead of availability. The test question is not how natural the arrangement is, but whether something unforeseen can still occur within it. The counter-operation is the maximisation of the capacity for contradiction, not of proximity or distance; the failure form would be a criterion of naturalness that disqualifies laboratories wholesale while overlooking that field observation, too, can predetermine its own answers.

**Conceptual stock.** Here there are two opposite traps: the hermeneutical gap, in which an experience finds no conceptual place, and pernicious abstraction, in which an accurate relational concept – "livestock," "pest," "population," "eco-point" – congeals into a description of essence. The first calls for supplementation, the second for de-absolutisation. The failure form is conceptual decree: one decrees a friendlier word without changing the measuring practices and relations through which the object actually becomes visible.

**Discourse pragmatics.** An objection can occur without being permitted to have effect: it is recorded, but classified as "personal impression," "sentimental," or "off-topic," thereby losing its status as a possible correction of the basis for decision. The counter-operation is making the accompanying presuppositions explicit and imposing a duty to justify rules of relevance. The failure form is counter-blockage, which in turn morally disqualifies every other position and thereby ends the dispute rather than opening it.

**The chain of translation.** Behaviour becomes findings, findings become prevalences, prevalences become an index of 72 points. Each step may be factually correct; the chain becomes de-addressing when it renders its own loss of differentiation invisible and the aggregate begins not merely to summarise the individual case but to represent it normatively. The counter-operation consists of return channels: anomalies must be able to compel a re-examination of the individual case. The failure form is articulatory inflation – more sensors, denser representation, weaker reference to what is represented.

**Practice and arrangement.** De-addressing is also embedded in walls, paths and divisions of labour: no one claims that the animal's welfare is meaningless, but no one stands any longer at a place from which the entire chain of action could be experienced. The counter-operation consists of reconnections, not the abolition of differentiation. The failure form is the panopticism trap: complete camera surveillance increases the density of control without expanding the animal's possibilities for action. The question is not whether everything is visible, but whether what is visible can change the practice.

**Institution.** Here the blockage is typically privative: there exists no effective channel of revision. The test question is what a body can trigger when it objects – and if the answer is "a statement," nothing has yet been gained. The counter-operation consists of procedural rights: associational standing to sue, ombudsman offices with a suspensive veto that does not preempt a decision on the merits but enforces a return path. The failure form is ritualised participation, which by its mere existence creates the impression that the interest is already represented.

**Theoretical form.** A theory can de-address by speaking very reasonably about animals: the difficulty of reality is translated into a neighbouring, solvable problem of argumentation. The test is the relation to anomaly – can an irritating case alter the basic categories, or is it merely redescribed until it fits the scheme again? The counter-operation is methodological critique; the failure form is hostility to theory, which raises the call for immediate experience without noticing that this call, too, is conceptually constituted.

**Extension to species and ecosystems.** Here the mode changes: it is not an existing experience that is represented indirectly, but a morally relevant addressee that must first be made visible, in mediated form, out of processes, relations and forms of loss. The double trap is called strict individualism (no sentient victim, hence no loss) and romantic holism (the ecosystem as a great subject). The counter-operation is the construction of chains of resonance – scientific sensing, legal facts, political mechanisms of revision – measured not by stability, but by generative stability.

**The meta-norm.** Running across everything stands Peirce's rule not to block the road of inquiry. It is necessary but not sufficient: a husbandry system can continually correct its models and yet still treat the animal exclusively as a bearer of quantified events. Revisability is the form of addressing, not its content. Its practical formulation reads: closure without finalisation – one may decide, as long as it remains recognisable under what conditions the decision must be reopened.

##### **3. What the passage yields**

The obvious objection to such a construction is that it amounts to a theoretical labyrinth: eight layers, three forms of blockage, a failure form for every repair, and at the end not a single directive for action. The objection hits upon something correct – the topology is not a decision procedure and does not replace normative justification. Its yield lies elsewhere.

First, **it decouples the question of status from the question of visibility**. Whoever wants to know whether a being counts morally must choose a normative theory. Whoever wants to know whether a practice de-addresses it can examine this largely independently of that choice. The test "Can what has been screened out return?" is equally applicable for utilitarians, Kantians and capability theorists alike – and it strikes its own position as well: the analysis of utilitarianism shows that a powerful operation of re-addressing can, in the next step, founder on complete commensurability.

Second, **it explains exclusion without an excluding conviction**. The widespread assumption that moral exclusion must be locatable somewhere as a sentence is empirically weak: most of those involved in the practices described do not consider animal suffering a matter of indifference. The topology offers a language for what nonetheless occurs, without inventing a culprit – and shifts the question from disposition to architecture.

Third, this yields a **distribution of the burden of proof that is operationalisable**. Whoever compresses, aggregates, maps or translates must show which differences are preserved, which are lost, and by what path lost ones can become relevant again. This is a demand that can actually be made of a certification system, an approval procedure, or an ethics committee.

Fourth, **applied ethics gains a subject matter of its own**. The usual picture – a principle is applied to a case – conceals precisely the interstices in which it is decided as which case something appears at all. The topology makes the architecture of these transitions itself the theme. This renders it transferable beyond animal ethics: wherever those affected cannot discursively control the interpretation of their situation – in care, in psychiatry, in paediatrics, in dementia, for future generations – the same eight places apply, albeit with different weightings.

Fifth, and this is the yield for the culture of argumentation: **criticism becomes testable rather than confessional**

## Layers of (De-)Addressing

 ![Pink pig in a modern barn and research space with a concrete floor, metal grating, and glass walls; it walks attentively to the right and lifts a foreleg.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_1_leib.png "Andreas Vieth: Body and Expression – the Animal as Living Body in Its Environment. Own image concept created with generative AI (2026).")

 ![Pig in a transparently depicted experimental course with cameras, sensor beams, floor markings, and two colour-coded possible movement paths.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_2_erhebung.png "Andreas Vieth: Data collection situation – between experimental setup and unexpected response. Own image idea created with generative AI (2026).")

 ![Pig with transparent information fields; prominently highlighted is "livestock," alongside "stock," "weight," and "performance," while terms such as "curiosity," "play," "bonding," and "fear" appear paler.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_3_begriffsvorrat.png "Andreas Vieth: Vocabulary – when an accurate category captures its object too completely. Own image concept designed with generative AI (2026).")

 ![A pig next to a diagram of a deliberation process; a note labelled "objection" leads into the discourse but is there diverted into the fields "personal," "off-topic," and "sentimental," never reaching the agenda.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_4_diskurs.png "Andreas Vieth: Discourse pragmatics — an objection is heard without being permitted to take effect. Original image concept created with generative AI (2026).")

 ![Pig in front of an infographic with the successive fields "Behaviour," "Finding," "Prevalence," and "Index 72"; a return path labelled "Re-examination of the individual case" leads back to the animal.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_5_%C3%BCbersetzung.png "Andreas Vieth: Translation chain – from individual behaviour to the aggregated index and back. Own image concept designed with generative AI (2026).")

 ![Transparent spatial representation of a production chain with a pig in the “Stable” area and separate stations “Transport,” “Processing,” “Packaging,” and “Product,” connected by paths and arrows.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_6_praxis.png "Andreas Vieth: Practice and Arrangement – the spatially and organisationally constructed separation of animal and product. Original image concept created with generative AI (2026).")

 ![Pig beside a schematic institution with a committee, documents, and a decision field; arrows connect "finding & material" to "ombudsperson office", "association action", and "revision", leading back into the procedure.](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_7_institution.png "Andreas Vieth: Institution – political representation through effective avenues for revision and correction. Own image concept designed with generative AI (2026).")

 ![Pig within a transparent geometric grid featuring the categories "preference," "cognition," and "reason"; snout and body parts protrude from the schema, connected to the fields "anomaly," "reinterpretation," and "methods critique."](https://andreasvieth.de/fileadmin/user_upload/slider_2_bild_8_theorieform.png "Andreas Vieth: Theoretical form – when the anomaly is either reabsorbed into the schema or taken as the occasion for a critique of method. Own image concept created with generative AI (2026).")

Body and Expression – the animal as a living body in its environment.

A pig moves attentively through a sober stable and research space. Even before measurements, categories, or institutional procedures come into play, it appears as a living body whose posture and movement are directed toward its environment. This layer thus holds fast to the point of origin to which later abstractions must be able to lead back.

Data collection situation – between experimental setup and unexpected response.

Cameras, sensor fields, floor markings, and two possible paths overlay the basic scene. The setting structures which reactions can become visible at all; at the same time, a deviation from the intended path remains possible. Addressing research reveals itself in the fact that the animal can still surprise the experimental setup.

Conceptual stock – when an accurate category grasps the object all too completely.

The pig is captured by a conceptual grid: “livestock,” “inventory,” “weight,” and “yield” stand out clearly, while “curiosity,” “play,” “attachment,” and “fear” are pushed to the margins. The categories are not necessarily false; they become problematic when they alone determine what remains visible and relevant about the animal.

Discourse pragmatics – an objection is heard without being permitted to become effective.

A documented objection concerning an animal's welfare enters a deliberative process, yet before reaching the agenda it is redirected into categories such as “personal,” “off-topic,” or “sentimental.” The blockage lies not in silence but between utterance and effect: the objection is registered without being able to alter the basis for decision.

Translation chain – from individual behaviour to aggregated index and back.

From concrete behaviour, a chain leads via finding and prevalence to a numerical index of 72. With each translation step, comparability increases, while particularities of the individual case may be lost. The feedback channel shown thus marks the possibility of returning from the aggregate to the individual animal.

Practice and Arrangement – the spatially and divisionally constructed separation of animal and product.

Stable, transport, processing, packaging, and product appear as spatially separated stations of a chain based on division of labour. Each station shows only a partial view; at the end, the product stands where the living animal is no longer present. De-addressing arises here through the architecture of practical separations.

Institution – political representation through effective channels of revision and correction.

Findings concerning the animal enter an institutional procedure of review and decision. However, an ombudsperson's office, associational standing to sue, and revision constitute pathways of return through which new evidence can reopen the decision. Representation here does not mean simulating a voice for the animal, but organising correction with real consequences.

Theory form – when the anomaly is either brought back into the schema or becomes the occasion for methodological critique.

A geometric schema is laid over the pig and orders it using categories such as preference, cognition, and reason. Snout, ear, and movement, however, exceed the grid and appear as an “anomaly.” The decisive alternative consists in either reinterpreting the anomaly to fit, or revising the categories of the theory itself.

##### Literature

**Bales, R. Eugene (1971).** Act-Utilitarianism: Account of Right-Making Characteristics or Decision-Making Procedure?, in: *American Philosophical Quarterly* 8 (3), pp. 257–265.

**Buytendijk, Frederik J. J. and Helmuth Plessner (1925).** Die Deutung des mimischen Ausdrucks. Ein Beitrag zur Lehre vom Bewusstsein des anderen Ichs, in: *Philosophischer Anzeiger* 1, pp. 72–126.

**Crary, Alice (2016).** *Inside Ethics: On the Demands of Moral Thought*. Cambridge, MA: Harvard University Press. Used in the project: in particular chap. 1, 3 and 4.

**Debaise, Didier (2017).** *Speculative Empiricism: Revisiting Whitehead*. Trans. by Tomas Joseph Weber, foreword by Isabelle Stengers. Edinburgh: Edinburgh University Press. Used in the project: chap. 3 "Creativity as Ultimate Principle" and chap. 4 "Actualising Creativity".

**Despret, Vinciane (2004).** The Body We Care For: Figures of Anthropo-zoo-genesis, in: *Body & Society* 10 (2–3), pp. 111–134. DOI: 10.1177/1357034X04042938.

**Diamond, Cora (1982).** Anything but Argument?, in: *Philosophical Investigations* 5 (1), pp. 23–41. DOI: 10.1111/j.1467-9205.1982.tb00532.x.

**Diamond, Cora (2003).** The Difficulty of Reality and the Difficulty of Philosophy, in: *Partial Answers: Journal of Literature and the History of Ideas* 1 (2), pp. 1–26. DOI: 10.1353/pan.0.0090.

**Donaldson, Sue and Will Kymlicka (2014).** Unruly Beasts: Animal Citizens and the Threat of Tyranny, in: *Canadian Journal of Political Science / Revue canadienne de science politique* 47 (1), pp. 23–45. DOI: 10.1017/S0008423914000195.

**Fisher, Alex (2025).** In Defence of Fictional Examples, in: *The Philosophical Quarterly*, advance online publication, 11 April 2025. DOI: 10.1093/pq/pqaf036.

**Fricker, Miranda (2007).** *Epistemic Injustice: Power and the Ethics of Knowing*. Oxford: Oxford University Press. Used in the project: chap. 7 "Hermeneutical Injustice", pp. 147–175.

**Krüger, Hans-Peter (2020).** *Homo absconditus: Helmuth Plessners Philosophische Anthropologie im Vergleich*. Berlin/Boston: De Gruyter. Used in the project: chap. 15 "Die Körper-Leib-Differenz von Personen: Exzentrische Positionalität und homo absconditus", pp. 393–408. DOI of the chapter: 10.1515/9783110666373-018. The PDF bears a copyright notice from 2019; the book was published in 2020.

**Kymlicka, Will and Sue Donaldson (2014).** Animals and the Frontiers of Citizenship, in: *Oxford Journal of Legal Studies* 34 (2), pp. 201–219. DOI: 10.1093/ojls/gqu001.

**Langton, Rae (2018).** Blocking as Counter-Speech, in: Daniel Fogal, Daniel W. Harris and Matt Moss (eds.), *New Work on Speech Acts*. Oxford: Oxford University Press, pp. 144–164. DOI: 10.1093/oso/9780198738831.003.0006.

**Latour, Bruno (2004).** How to Talk About the Body? The Normative Dimension of Science Studies, in: *Body & Society* 10 (2–3), pp. 205–229. DOI: 10.1177/1357034X04042943.

**Luhmann, Niklas (1990).** *Ökologische Kommunikation. Kann die moderne Gesellschaft sich auf ökologische Gefährdungen einstellen?* 3rd ed., Opladen: Westdeutscher Verlag. Used in the project: in particular chap. XII "Wissenschaft", XIII "Politik" and XVII "Beschränkung und Verstärkung: Zu wenig und zu viel Resonanz".

**McLaughlin, Amy L. (2014).**

#### Addressing and De-Addressing of Animals, Animal Species and Ecosystems

## 01 | The Fundamental Problem of Epistemic Asymmetry

**How can we do justice to subjects or systems that are unable to express their own interests in propositional justifications?**

Animals are not devoid of expression. They flee, seek, avoid, refuse, cooperate, resist, alter their behaviour, and react to changes in their environment. A dog can avoid a situation, a cow can struggle against being restrained, a laboratory animal can refuse a task. None of this already constitutes an argument. The animal does not say: *This treatment is wrong, and for the following reasons.* But it does not follow that its behaviour is epistemically meaningless. It can place our descriptions under **evidential pressure**: what we assert about an animal, its needs, or a situation must prove itself against how the animal actually lives, acts, and reacts.

This is precisely where the fundamental problem of *Experience without Voice* begins. Whoever speaks about non-human beings finds themselves in an asymmetrical position. Human beings possess the concepts, write research reports, determine experimental designs, formulate laws, and decide which observations count as relevant. The animal cannot participate in these discursive procedures in the same way. It cannot answer an interpretation with a counter-expert opinion. This asymmetry does not diminish its moral significance. Rather, it changes the demands placed on those who describe and represent it. Where the possibility of propositional self-representation is absent, the **methodological burden of scrutiny** increases.

This is meant to avoid a twofold error. The first consists in an **anthropocentric silencing**. It equates discursive linguistic capacity with the capacity for expression, and thereby readily treats non-linguistic living beings as passive objects. Behaviour then appears merely as a reflex, a measured value, or the output of a mechanism. Ludwig Wittgenstein points in a different direction. For him, our concept of pain does not depend first on a theoretical inference from one's own inner life to a foreign inner life. We react to pain-behaviour; the turning towards the suffering other belongs to a pre-linguistic practice on which later language-games are built in the first place. His famous juxtaposition of a stone and a “buzzing fly” makes clear that our attitude towards the living is already structured by its manner of appearance and behaviour. In the *Zettel*, Wittgenstein calls the tending of the other's painful spot “primitive” for precisely this reason: not because it is mentally primitive, but because it is pre-linguistic and carries a way of thinking rather than being derived from one.

A related point can be found in Buytendijk and Plessner. Movements such as grasping, fleeing, warding off, or searching do not first confront us as a sum of mechanical individual movements from which we subsequently compute a meaning. They can appear as **gestalts of movement**: as bodily directed behaviour of a living being towards its environment. Only a particular analytical abstraction breaks this gestalt down into joint angles, muscle movements, or kinematic sequences. This guards against a first form of de-addressing: an animal is not epistemically invisible simply because it cannot formulate its state in sentences.

The opposite error, however, would be equally problematic. It consists in a **naive paternalism of advocacy**. Whoever takes an animal seriously must not simply translate its behaviour into human sentences: *It is telling us that it wants freedom*; *it is protesting against animal experimentation*; *it is demanding a right to self-determination*. Such formulations may be rhetorically effective, but they blur a decisive difference. An animal that flees or resists being restrained performs an action; it does not thereby already raise a morally grounded objection. The human interpretation must not replace the missing propositional voice with an imagined one.

For this reason, *Experience without Voice* distinguishes three things that are easily conflated: **bodily expression, epistemic correction, and moral objection**. Bodily expression occurs on the side of the animal. Epistemic correction arises where this behaviour becomes relevant to our description. And a moral objection already belongs to a normative language of reasons, rights, and duties. An animal therefore does not “refute” our theory in the logical sense. But its behaviour can make a theory look bad. It can compel an explanation, a husbandry model, or an experimental design to change. It is precisely this intermediate position that the term **evidential pressure** designates here.

For this idea, the works of Vinciane Despret and Bruno Latour are especially illuminating. Despret's analysis of “Clever Hans” shows how strongly scientific knowledge depends on which relations an experimental design allows to become visible at all. The horse could not solve mathematical problems; but it could perceive minimal bodily signals from the humans that they themselves did not notice. The experiment thus did not merely refute a false attribution of arithmetical ability. It made a previously invisible relation between human and animal bodies investigable. Latour draws a general methodological conclusion from this: to have a body means learning to perceive differences and to let oneself be affected by them. Knowledge does not necessarily improve by removing the body from the situation, but by making the situation articulate ever finer differences.

This also shifts the picture of scientific objectivity. A good description is not simply one in which the observer disappears as completely as possible. What matters, rather, is whether an investigation is constructed in such a way that differences can make a difference. An experimental design in which an animal always confirms the same theoretical interpretation regardless of its behaviour is epistemically suspect. It has blocked the path back from experience into theory.

Here the project touches upon Charles Sanders Peirce's famous rule: **“Do not block the way of inquiry.”** For Peirce, knowledge requires the willingness to remain dissatisfied with one's own current state of belief and to use theories in such a way that further inquiry remains possible. Especially problematic for him are claims that something is ultimately unknowable, absolutely inexplicable, or already conclusively formulated. Amy McLaughlin accordingly highlights the fallibilist core of Peirce's rule: research possesses a self-correcting structure only if it actually remains open to learning. For *Experience without Voice*, a simple methodological rule can be derived from this: **a description of the other must not be constructed in such a way that nothing the other does can still disturb it.**

This rule is especially important because an **epistemic asymmetry** exists here. Miranda Fricker describes, for human social relations, how power can determine which experiences become comprehensible and articulable at all. Hermeneutical injustice arises where people, owing to structurally distorted interpretive resources, have difficulty putting into words experiences whose understanding matters essentially to them. The situation of non-human animals is not simply a case of Fricker's hermeneutical injustice: animals are not merely speakers who lack a suitable concept. But the analogy makes something visible. Here too, the available categories are supplied by the more powerful side. Precisely for this reason, these categories must not be confused with the object they are meant to capture.

The consequence is not a demand for epistemic abstinence. We cannot stop describing animals merely because every description is perspectival and fallible. We must describe them when we treat them medically, keep them, protect them, research them, or alter their habitats. The consequence, rather, is **triangulation**. An attribution should remain checkable, as far as possible, against several mutually distinguishable forms of evidence: behaviour and expression, physiological findings, the concrete situation, comparative behavioural research, and the possibility of alternative interpretations. The smaller the possibility for the party concerned to correct the human interpretation discursively itself, the greater becomes the duty to keep such avenues of correction methodically open.

In this context, a second conflation of levels must also be avoided. **Experience, moral relevance, and practical decisiveness are not the same thing.** That an animal feels pain is, in the first instance, a statement about how it is faring. Ethics does not generate this pain. That this pain constitutes a moral reason is already a normative question. And even if it is a moral reason, it does not follow, without further premises, what must be done in a concrete situation. Moral relevance generates a need for justification; it does not yet guarantee practical decisiveness.

This control of transitions is central to the project. It prevents two opposite shortcuts. One blinds itself to experience until moral theory appears to operate only with abstract principles. The other transforms every experienced ill immediately into an unambiguous duty. With her notion that humans and animals are “inside ethics,” Alice Crary opposes a picture according to which the empirically knowable world is initially wholly morally neutral and value is only added afterwards. For her, human and animal forms of life can be perceived in a way that is already morally significant. The project *Experience without Voice* adopts from this above all a warning: one must not narrow the supposedly “neutral” description in advance to such an extent that the morally relevant can no longer appear in it at all. At the same time, it adheres more strongly than Crary to an analytical control of transitions: from the perception of moral significance, a concrete duty should not follow without further justification.

A third distinction concerns the concept of **agency**. Animals can act, learn, form expectations, shape relationships, and respond to social norms without thereby being *moral agents* in the strong sense. Donaldson and Kymlicka explicitly oppose an over-intellectualized concept of social and political agency. Precisely everyday forms of cooperation, consideration, self-restraint, and sensitivity to norms rest, even in humans, by no means exclusively on explicit rational deliberation and assent to moral propositions. Many social animals likewise possess corresponding prosocial dispositions and capacities. This does not thereby make them participants in a philosophical discourse of justification. But it suffices to reject the alternative “either rational moral agent or passive object.”

This also clarifies further what **representation** can mean here. Human ethics does not first grant tierlicher \[animal\] experience its being. Nor does it generate the agency it describes. Its task, rather, consists in **asserting** this experience and capacity for action within human epistemic, moral, and political spaces. Precisely because this assertion cannot simply be authorized, linguistically specified, and, where appropriate, revoked by the animal, it remains accountable. It must disclose which observations it uses, which concepts it employs, which alternative interpretations are possible, and at which point it passes from an empirical description to a normative evaluation.

For ecological entities, the problem shifts once again. An ecosystem, a population, or a biotope does not possess bodily expression or experience in the same sense as an animal. For this reason, one should also refrain here from prematurely speaking metaphorically of a “voice of nature.” Nevertheless, ecological contexts can generate **evidential pressure**: models fail, expected stabilities collapse, interactions occur that a theory had not anticipated. Here too, then, the Peircean rule applies: not to block the way of inquiry by allowing our categories to register only what they already expect. The commonality between animal and ecosystem thus does not lie in both “speaking,” but in the fact that a responsible description must keep open **avenues for disturbance and correction**.

The fundamental principle of *Experience without Voice* can therefore be summarized in one sentence: **where propositional self-representation is absent, representation must not become freer, but methodologically more demanding.** Non-linguistic beings and non-discursive systems are done justice neither by silencing them nor by putting our own language into their mouths. What is decisive is whether our descriptions are constructed in such a way that the other can still make a difference within them.

At this point, the term evidential pressure

## Sensory Modes Before Language: Bodily Expression and Environmental Intentionality

**At which level does the perception of animal experience begin, before scientific models or moral concepts take hold?**

Before an animal becomes the object of a distress scale, an experimental result, or a moral argument, it already encounters us as a living body within an environment. It searches, flees, pauses, evades, approaches, threatens, or protects an injured part of its body. Such processes are not yet moral arguments. But neither are they, in the first instance, meaningless bodily movements to which we subsequently add meaning. This first layer of *Experience without a Voice* is therefore concerned with an elementary form of addressing: **Living behaviour appears as the directed behaviour of a body within an environment.**

"Prior to language" does not mean: prior to any conceptual shaping, in a wholly theory-free or infallible form of perception. What is meant is something more modest, and at the same time more fundamental. Before behaviour becomes propositional self-description, scientific modelling, or moral justification, we can already perceive that an animal is searching, fleeing, or evading. A dog need not be able to say *"I am searching for something"* for its behaviour to appear as searching. Nor must we first measure heart rate, movement data, or neural activity in order to perceive this searching.

Precisely this initial mode of access may later be corrected, differentiated, and expanded through empirical research. But it should not disappear from view merely because it is not yet a scientific explanation.

For this layer, Frederik J. J. Buytendijk and Helmuth Plessner, in their 1925 work *Die Deutung des mimischen Ausdrucks* \[The Interpretation of Mimetic Expression\], provide a particularly precise model. Their starting point is not a visible body behind which a hidden psychic inner life would have to be inferred. The starting point is rather the **behaviour of a living organism within its environment**. Movements such as walking, grasping, fleeing, warding off, or searching appear as temporally extended movement-gestalts. Whoever perceives them does not first see individual muscle contractions and only then calculate that this is presumably "flight" or "searching". The movement already possesses a gestalt directed towards the environment.

Hans-Peter Krüger captures this point with Plessner's concept of **environmental intentionality**. Living beings are corporeal insofar as their behaviour enacts a directedness of meaning towards their environment. This directedness is not a physical property like mass, velocity, or joint angle. Yet it can be perceived and understood in the appearance of behaviour itself. The same occurrence can simultaneously be investigated in bodily-functional or physiological terms. Leib \[lived body\] and Körper \[physical body\] therefore do not designate two different things, but different modes of access to the same living being.

Buytendijk and Plessner illustrate this with a simple example: a dog runs back and forth with lowered head, sniffs, stops, turns back, and changes direction. In this sequence of movements we see **searching**. The word is not a poetic embellishment of an otherwise meaningless accumulation of changes in location. It designates the gestalt which these movements assume precisely within the relation between animal and surroundings.

And yet the same occurrence can be grasped differently. Movement can be decomposed into distances, speeds, joint positions, stops, and changes of direction. This analysis is neither false nor philosophically suspect. It brings other properties of the same process into view. What interests Buytendijk and Plessner is the point at which analytical decomposition does not merely supplement the original gestalt but replaces it. Once the organism is detached from its relation to the environment, what remains is a measurable sequence of movements; the context in which this movement *is* searching can thereby be lost. For this they memorably speak of a **"dead melody"**.

**Three levels of the same occurrence can thus be distinguished:** First, we encounter an animal that behaves within a situation. The same occurrence can then be analysed in bodily and physiological terms. And it can be understood as a lived behavioural gestalt whose meaning arises from the relation between animal and environment. None of these levels need eliminate the others. What matters, rather, is whether the transition between them remains open.

The philosophical point therefore does not lie in an alternative between "science" and "phenomenology". A physiological description can be highly precise. Muscle activity, neural processes, hormones, or movement trajectories can reveal important differences that mere observation would miss. The analytical approach becomes problematic only when it forgets its own act of abstraction and declares itself the complete description of the living being.

For *Experience without a Voice*, this constitutes a first site of possible **de-addressing**. It does not arise simply because behaviour is measured. It arises where methodological reduction turns into ontological reductionism: an animal that does something then becomes a body to which something merely happens.

This can be captured through a simple distinction. For a particular research purpose, it may be sensible to measure the speed of a movement, a heart rate, or the frequency of a behaviour. It does not follow, however, that the behaviour *actually* consists only of these parameters. Measurement abstracts from certain properties of the occurrence in order to determine others more precisely. Its strength lies precisely in this restriction. It becomes de-addressing only once what has been excluded is in principle no longer permitted to return.

This also clarifies why **environmental intentionality** does not designate some mysterious inner realm of the animal. When we say that a dog is searching, we are not claiming that behind its visible movements lies an invisible mental object called "searching". The form of meaning lies in the relation itself: the behaviour receives its gestalt from the context of organism, situation, possible goals, obstacles, and changes in its environment.

For this reason, meaning also does not reside in an isolated signal. A raised head is, by itself, neither attentiveness nor threat. Rapid running is not necessarily flight. Ruffled feathers, retreating, or fixating acquire their meaning only within a temporal and situational context. **Meaning lies in the behavioural gestalt, not in a single data point.**

This is precisely what also guards against the opposite error, that of **unexamined anthropomorphism**. If a mechanistic description leaves too little meaning to an animal, it would not be an appropriate correction to attribute to it, in return, arbitrary human ideas and justifications. A dog searching for a scent trail need not imagine the same thing as a human searching for their keys. An animal that threatens does not thereby formulate a theory of social sanction. And an animal that resists a treatment does not thereby establish a right to bodily self-determination.

The layer of expression therefore moves between two failures: **we need not empty behaviour of meaning in order to avoid anthropomorphism; but neither need we humanise it in order to perceive its meaning.**

The appropriate description remains species-specific, situation-dependent, and revisable. "Searching", "fleeing", "avoiding", "warding off", or "grasping" can be precise descriptions even though the same words are also used for human behaviour. What matters is not their linguistic origin but whether the designated gestalt holds within the concrete relation of body and environment. If observations speak against it, the description must remain open to revision. Here begins already the **pressure of evidence** that runs through the first accordion: expression is not an infallible revelation of a hidden interior, but it is a form of evidence against which our descriptions can fail.

Ludwig Wittgenstein's reflections on pain develop the same thought from a different direction. In the *Philosophical Investigations*, he asks how we come to attribute sensations to another being at all. His answer does not proceed by way of an inference to a hidden interior. Wittgenstein contrasts a stone with a "fly struggling" \[zappelnde Fliege\]. In the case of the stone, the idea that it has sensations seems peculiarly unfounded; in the case of the struggling fly, this difficulty disappears. "Our attitude to what is living," he writes, is not the attitude to what is dead. Nor does describing a facial expression consist in stating the measurements of a face.

This does not mean that certain visible movements **prove** pain. Wittgenstein's point runs deeper. Our very practice of speaking of pain, sensation, and suffering beings is already embedded in ways of reacting to the living. We do not begin with neutral movement data and subsequently perform a metaphysical inference by analogy to another's consciousness. Concepts of pain live within a context of expression, attention, care, comfort, and our dealings with vulnerable beings.

In the *Zettel*, Wittgenstein therefore calls it "primitive behaviour" to tend to another's aching spot and to attend to their pain-behaviour. "Primitive" here expressly means: **pre-linguistic**. The language-game is built upon this behaviour; it is the "prototype of a way of thinking" and not the result of an inference. The explanation that we care for others because we have inferred, by analogy with our own case, their experience of pain, is accordingly called by Wittgenstein "wrongly harnessed".

This gives the title **"Forms of Meaning prior to Language"** its precise sense. Animals do not carry their inner states transparently on the outside. Nor is our perception infallible or independent of experience and concepts. But propositional justification does not stand at the beginning. Even before we can explain why a particular behaviour counts as pain-behaviour, flight-behaviour, or searching-behaviour, we already move within practical distinctions between the living and the dead, approach and avoidance, turning towards and withdrawing.

Wittgenstein's remarks should therefore not be read as an alternative "pain-detector theory". His examples examine the **grammar and form of life** within which concepts of sensation have meaning at all. For *Experience without a Voice*, their point lies in the fact that expression, and our reaction to expression, belong to the very concept of a sentient being.

This allows the structure of blockage in this first layer to be specified more precisely. De-addressing does not first occur where someone explicitly denies that animals can suffer. It can begin far less spectacularly: when an institution registers only what appears within its measurement format. A cow that repeatedly withdraws from the feeding rail becomes a reduced feed intake; a mouse that systematically explores the edge of a test enclosure becomes a movement trajectory; a dog that avoids a situation becomes an altered probability of presence. All of these data can be correct and important. **They become de-addressing only when the gain in information through measurement simultaneously blocks the way back to the occurrence of which it is the measurement.**

The counter-operation is therefore not to **abolish data**, but to **keep expression retrievable**. Physiological, behavioural-scientific, and quantitative methods can considerably improve our perception. Yet they must remain tied back to the behaviour from which their variables were abstracted. A heart-rate curve says something different from a flight attempt; a trajectory something different from searching. Sound knowledge does not require choosing between these languages, but keeping the transitions between them controllable.

Understood in this way, Layer 1 forms the phenomenal foundation of the entire model. Later layers may ask what experience may be attributed to an animal, what scientific statements follow from this, what moral relevance this experience possesses, and what duties arise from it. But they must not retroactively eliminate this starting point.

Before the animal becomes the case of a theory, it is encountered as a body that behaves within an environment.

#### Visualization of the Three Levels (Buytendijk/Plessner)

 ![The same recording with captions indicating environmental reference, search movement, change of direction, stopping, and scent trail.](https://andreasvieth.de/fileadmin/_processed_/7/9/csm_hund__3__bd5a98c93c.png "Andreas Vieth: Phenomenological view: behavioral gestalt of the searching movement. Own image concept created with generative AI (2026).") ![Brown-and-white spotted short-haired pointer wearing a harness on a field path, head lowered close to the ground, sniffing; tall grass and herbs on the right, meadows and wooded hills in the background.](https://andreasvieth.de/fileadmin/_processed_/d/4/csm_hund__1__2b1549105a.png "Andreas Vieth: Baseline recording without superimposition: dog searching in the field. Own visual concept created with generative AI (2026)")

 ![Same footage with skeleton tracking, motion trajectory, and diagrams of joint angles, ground contact, and velocity](https://andreasvieth.de/fileadmin/_processed_/7/e/csm_hund__2__a7e2af913b.png "Andreas Vieth: Physiological-scientific view: motion analysis through skeletal tracking. Own image concept designed with generative AI (2026).")

- Andreas Vieth: Physiological-scientific view: motion analysis through skeletal tracking. Own image concept designed with generative AI (2026).
- Andreas Vieth: Baseline recording without superimposition: dog searching in the field. Own visual concept created with generative AI (2026)
- Andreas Vieth: Phenomenological view: behavioral gestalt of the searching movement. Own image concept created with generative AI (2026).

**Slide to the left or to the right:** In the **middle**, you initially see the dog without any analytical overlay. Slid to the **left**, the **physiological analysis** of the dog appears – movement as distance covered, speed, joint position and change of direction. Pulling the slider to the **right** makes its movement visible as a **behavioural gestalt**: the dog sniffs, pauses, turns around – it is **searching**. It is precisely this distinction that interests Buytendijk and Plessner. The physiological analysis is not false; it renders other properties of the same occurrence visible. It becomes problematic only when the decomposition replaces the relation between body and environment. What then remains is a measurable sequence of movements, while the context in which this movement constitutes *searching* may be lost – a **"dead melody."**

## 03 | Data Collection Situation & Setting: Docility vs. Availability

How do our measurement setups, laboratories and observation methods influence what an animal is able to show at all?

Scientific knowledge about animals does not arise before a neutral stage. It arises in concrete **data-collection situations**: in the laboratory, in the maze, before a camera, in a behavioural test, or in a particular encounter between animal and researchers. These situations render some behaviour visible and other behaviour invisible. They can open up possibilities for an animal to answer a question in an unexpected way – or they can already fix the possible answers through the experimental setup itself.

This shifts the problem of de-addressing. The first layer concerned **how animal behaviour is perceived and described**. Now we take a step back: **Under what conditions could this behaviour emerge at all?** An animal cannot simply “show what it can do” within a scientific investigation. What becomes visible also depends on which tasks are set, which apparatuses are built, and which possibilities of response are permitted.

Vinciane Despret brings out this relational structure of scientific situations with particular clarity. Her most famous example is **Clever Hans**. At the beginning of the twentieth century, the horse Hans appeared able to solve arithmetic problems by giving the correct number of taps with his hoof. Oskar Pfungst eventually showed that Hans was not calculating. But for Despret, this is precisely not where the interesting discovery ends. Hans was reacting to tiny, involuntary movements of the questioning humans: to changes in their posture and tension of which they themselves were entirely unaware. He was able to read human bodies – and thereby made visible how much these bodies expressed without knowing it. The experiment therefore did not simply reveal an “error” on the horse's part. Rather, Hans became an instrument through which human beings could learn something about their own bodily existence.

This shift becomes even clearer in Despret's reconstruction of the **Rosenthal experiment**. Students were given ordinary laboratory rats that had arbitrarily been labelled “bright” or “dull.” Nevertheless, differences appeared in the experimental results. Among other things, the supposedly gifted animals were handled more gently by the students and perceived as more pleasant. Rosenthal interpreted such effects chiefly as a disruptive influence of the experimenters' expectations, which good research must, as far as possible, control for.

Despret, however, takes the point a step further still. The situation need not be understood exclusively as a distortion of an otherwise pure reality. Students and rats **altered one another**. Expectations could lead researchers to treat an animal more attentively, more carefully, and with greater interest; animals, in turn, could respond to this altered relationship. Despret therefore speaks of a reciprocal **availability**: researchers make themselves available for the behaviour of the animal, and animals become available for new relationships and capacities. This is something other than the idea of a neutral researcher unchangingly reading off data from an unchanging object.

What matters here is a distinction that Despret herself explicitly introduces: **availability is not docility**. An animal is not epistemically interesting because it does precisely what the experimenter expects. Availability shows itself, rather, in the fact that something not already settled can still happen within the experimental arrangement. Despret reports, for instance, on a rat classified as “dull” that proved, contrary to expectation, to be exceptionally capable. Precisely because the animal did not simply fulfil the role assigned to it, it was able to add something to the situation.

The counterpart, for Despret, is Harry Harlow's research with isolated rhesus monkeys. Harlow separated young monkeys from their mothers and peers and constructed experimental arrangements meant to investigate the consequences of social isolation. For Despret, the problem does not lie only in the cruelty of these experiments. It is at the same time an **epistemological** problem. The apparatus was set up such that despair was precisely the result it was designed to produce. How could a monkey have objected to this experimental arrangement? Despair would only have confirmed it. The setting could hardly open up any other form of life to it. The animal was articulated by the apparatus, instead of itself participating in the articulation of the research question.

Here lies the difference between **docility** and **availability**. Docility does not simply mean that an animal behaves calmly or can be trained. What is meant is an epistemic structure: the experimental setup leaves its object only the possibility of reproducing the presuppositions of the experiment. Availability, by contrast, designates a situation in which the participants – animal, researchers, and apparatus – can themselves be changed by what occurs. For Despret, this is explicitly not a sentimental demand for special affection towards animals, but a condition of more interesting science: good questions make **articulated answers** possible, and articulated answers can, in turn, change the questions.

Bruno Latour radicalises this thought with reference to Despret and Isabelle Stengers. For him, an experimental arrangement is not scientifically interesting merely because a hypothesis is theoretically falsifiable. The real risk begins only where **the question itself, the research programme, the protocol, and the apparatus** can be called into question by the behaviour of the being under investigation. Researchers should therefore not merely ask: *Does the animal answer my hypothesis with yes or no?* They must also ask: *Have I even posed the right question? Is the animal responding to the phenomenon I wish to investigate – or to an artefact of my experimental setup?*

Latour calls this capacity for objection **recalcitrance**. The word does not designate a character trait of a stubborn animal. What is meant is the possibility for the object of investigation to withdraw from the scientific ascription made of it. A good data-collection situation therefore does not maximise obedience, but rather **capacity for objection**. It gives the animal the opportunity not to fulfil the researchers' categories, and thereby to make visible that something about the question, the apparatus, or the interpretation needs to be changed. Latour formulates this pointedly as a methodological principle: investigations should be arranged so that the recalcitrance of those questioned can become as great as possible.

This allows the form of de-addressing at this second layer to be precisely determined. It does not consist in researchers measuring, testing, or subjecting animals to experimental conditions as such. Even artificial apparatuses can be extraordinarily rich in what they reveal. The blockage arises, rather, when the data-collection situation **immunises its own presuppositions against correction**. The animal then appears to be answering, but it can only answer within an already fixed repertoire. The method produces its own confirmation.

The decisive question is therefore not: **How natural is the setting?** Even a laboratory can be an excellent site of scientific knowledge. The better question is: **Can something still happen in this setting that the researchers did not foresee?** A good experiment does not aim to produce as little relationship as possible between researchers and animal. It organises this relationship so that surprise, resistance, and revision remain possible. Distance or empathy are, in themselves, not criteria of quality. What matters is whether the encounter expands the repertoire of what animal and researchers can subsequently say to one another.

Thus the second layer of de-addressing obtains its own test mark: **a setting de-addresses an animal precisely where it blocks the way back for the unexpected.** It allows data to arise, but prevents the animal from calling into question the conditions of their emergence. Addressing research, by contrast, keeps this way back open. The animal need not be able to speak in order to object scientifically. It suffices that the experimental arrangement leaves it the possibility of answering differently than we expected.

## 04 | Conceptual Stock & Hermeneutics: Distorting Categories and Conceptual Invisibility

Why does the understanding of animal experience or ecological relationships sometimes already fail at the level of the very concepts with which we speak about them?

After the first two layers asked **how animal expression is perceived** and **under what conditions an animal can show something at all**, the third layer takes yet another approach. Even once a behaviour has become visible and has been scientifically adequately recorded, it must subsequently be **conceptually grasped**. It enters an existing stock of words, categories and patterns of interpretation. This conceptual stock is not a neutral wrapping of an already fully cognised reality. It co-determines which differences we are able to register at all, which descriptions appear self-evident – and what withdraws itself from our shared understanding.

De-addressing can occur at this level in two almost opposite ways. Sometimes a **concept is missing**: an experience or a relationship cannot be adequately articulated with the available categories. Sometimes, however, there is by no means a shortage of concepts. Rather, a concept is already in place that organises the situation so effectively that other descriptions can scarcely come into play any longer. The first form is **privative**: something is missing. The second is **preclusive**: something already present occupies the conceptual space.

For the first form, Miranda Fricker's theory of **hermeneutical injustice** offers an important model. Fricker examines situations in which people cannot adequately understand or convey to others essential areas of their own social experience, because a gap exists in the collectively available interpretive resources. Her well-known example is the emergence of the concept of "sexual harassment". Before such a concept was available, concrete instances of assault could certainly be experienced – yet those affected had no shared category under which different experiences could be recognised, communicated and criticised as belonging together. The problem, then, did not consist in nothing happening. It consisted in the fact that what happened had no adequate place in the shared conceptual stock.

However, Fricker is speaking of human speakers who, owing to social power relations, are unequally involved in the production of shared interpretive resources. Her concept of **hermeneutical injustice** therefore cannot be transferred unaltered to animals or, indeed, to ecosystems. Animals do not, in Fricker's sense, attempt to render their experience conceptually intelligible within public discourse; ecosystems are not experiencing speakers at all. For animal and environmental ethics, her analysis is therefore initially a **hermeneutical analogy**. It directs attention to the question of what happens when the human resources of translation prove insufficient for animal experience or for ecological processes and forms of loss.

In the case of animals, this asymmetry is especially consequential. An animal can suffer, seek closeness, form bonds, lose a fellow creature, or react to a lasting burden, without itself supplying the concept under which this occurrence is to be socially understood. Its experience is present, but it is not **discursively self-representing**. The human linguistic community must translate expression, behaviour, corporeality and scientific knowledge into conceptual descriptions. This may be called **indirect addressing**: the animal has experience but cannot itself assume the public conceptual representation of that experience. In the case of ecosystems, the task shifts once more. Here it is not an existing experience that must be translated; rather, a morally relevant addressee must first be made visible, in a mediated way, out of processes, relations, vulnerabilities and forms of loss.

The privative danger is therefore this: **what happens when the appropriate categories are lacking?** Behaviour can then indeed be observed, yet it remains conceptually underdetermined. Long before, for instance, differentiated descriptions of animal social relationships, grief reactions, or states of chronic stress become available, the corresponding phenomena may already occur. Whatever lacks a sufficiently stable conceptual place does not thereby vanish from reality. Rather, it risks becoming unable to connect with social and moral understanding.

Yet conceptual de-addressing does not arise solely from such **gaps**. It can be especially effective precisely where a situation already appears to have been named quite unambiguously. Here William James's critique of **vicious abstractionism** is helpful. James does not thereby designate abstraction as such as an error. Without abstraction, science, administration and practical thought would be impossible. The error begins where an accurate feature is extracted from a complex, concrete situation and is subsequently used as though the object were **nothing other than** what is captured under this feature. The abstraction then does not simply add a perspective to the description – it practically effaces the rest.

This can be shown particularly well by the concept of **"livestock" \[Nutztier\]**. A pig can indeed be livestock. The designation is, in this respect, neither false nor necessarily disparaging. Strictly speaking, however, it initially designates **a relation**: human beings keep this animal within a certain practice for certain purposes. "Livestock" thus does not describe the essence of the pig, but its position within a relationship established by human beings.

Vicious abstraction sets in only once this accurate description of a relation unnoticeably becomes a complete description. Then "This pig stands to us in a relation of utilisation" practically becomes: **"What is relevant about this pig is what is relevant within this relation of utilisation."** Body weight, feed conversion, reproductive performance, disease susceptibility or mortality become visible at high resolution. Play behaviour, bonds, boredom, fear or individual peculiarities can simultaneously drop out of the relevant space of description. The animal is by no means poorly observed. On the contrary: it can be intensively measured and scientifically described – only under a category that co-determines which of its properties count as worth knowing.

The same pattern can be observed with the term **"pest" \[Schädling\]**. This concept, too, can initially express a real relation: a particular animal species stands, under given conditions, in conflict with human agricultural, forestry-related, or health-related goals. It is not de-addressing merely to name this conflict. The category becomes problematic once the relation congeals into an explanation of essence. "This animal damages our crops under these conditions" then becomes "This animal is a pest." Its ecological function, its own vital activities, or the question of through which human interventions the conflict arose in the first place, recede behind the category of control and eradication.

Something similar holds for technical terms of environmental administration. A forest can be described as **timber stock**, a moor as **area**, biodiversity as **species count**, an ecosystem as a **service**, or an intervention as a number of **eco-points**. All these abstractions can be meaningful, and even indispensable, for particular purposes. The error does not lie in the number, the function, or the administrative category. It begins where the grid causes one to forget that it is a grid. A historically evolved, process-like habitat then becomes a calculable quantity, and the question of its concrete, particular structure, or of irreversible losses, scarcely appears any longer within administrative language. Nature can be measured and administered in great detail and nevertheless be de-addressed precisely through **resource language, administrative grids and compensation logic**.

Law, too, shows how ambivalent such categories are. Under German civil law, animals are explicitly **not things**; at the same time, the provisions applying to things are largely applied to them correspondingly. The conceptual difference is thus present, without this making all structures of ownership, availability and interchangeability disappear. Precisely for this reason, it would be too simple to pin de-addressing on a single word. What matters is not whether a statute says "animal" or "thing", or an agricultural statistic says "animal welfare" or "performance". What matters is **which descriptions become practically relevant through a categorial order, and which systematically find no way back into perception**.

This also shows why the remedy cannot simply consist in exchanging supposedly bad words for supposedly good ones. Whoever replaces "livestock" with a friendlier expression has not thereby altered any relation of utilisation. Whoever says "animal welfare" may continue to capture exclusively those parameters that prevent losses in production. And a "compensation area" does not become ecologically equivalent merely because the administration designates it as such. The faulty form of the counter-operation may therefore be called a **conceptual decree**: one linguistically prescribes how something is to be seen, without altering the categories, measurement practices and institutional relations through which it is actually made visible.

Re-addressing demands something more exacting. In the case of a **privative blockage**, missing distinctions must be developed: we need richer descriptions wherever existing concepts swallow up differences. In the case of a **preclusive blockage**, existing categories need not necessarily be abolished, but rather **de-absolutised**. "Livestock", "pest", "invasive species", "stock" or "eco-point" may remain legitimate concepts in certain contexts. They must simply not be granted the right to decide, on their own, what the designated object is and what may count with regard to it.

Herein lies the particular point of this third layer. **De-addressing does not begin only with silence about a matter. It can also arise through an extraordinarily successful description.** A category can be correct and nevertheless gain too much power. The methodological question is therefore not merely: *Do we have a concept for this?* It is equally: **What does this concept let appear – and what does it no longer allow to return?**

The two forms of conceptual blockage can thus be briefly distinguished: in the case of the **hermeneutical gap**, there is something that our concepts cannot yet adequately capture. In the case of **vicious abstraction**, there is a concept that functions so well that it displaces other descriptions. In the first case, a path into shared language is missing. In the second, the path exists – but, owing to a dominant category, can be travelled in only one direction.

## Discourse Pragmatics & Blocking: The Preclusive Deflection of Objections

How can an objection disappear even though no one has refuted it?

With the fourth layer, the perspective shifts once again. At the level of **conceptual repertoire**, the question was whether an experience could be adequately described at all with the available categories. Now it is no longer primarily description but the **situation of speaking and objecting** that stands at the centre. An argument can be clearly formulated, empirically grounded and relevant to the matter at hand – and yet exert no effect in the further course of events. It is heard, perhaps even recorded, but treated in such a way that it can no longer alter the decision underway.

Discourses therefore do not consist solely of assertions, reasons and refutations. They have a **pragmatics**. There are roles, responsibilities, agendas, established burdens of proof and background assumptions about what may count as objective, scientific or procedurally relevant. Such rules are unavoidable: without them, no commission could deliberate and no scholarly dispute could be conducted with focus. De-addressing begins where this order not only reduces complexity but **devalues certain objections before they have even been examined on their merits**. An argument then fails not because a better argument has been given, but because of its pragmatic classification: *sentimental*, *unscientific*, *personal*, *off-topic*, or *not the subject of this procedure*.

Rae Langton's philosophy of language helps us understand how such an effect can come about. Her starting idea is initially unspectacular: in conversations, by no means everything is explicitly asserted. Utterances carry along presuppositions that remain in the background. Langton calls particularly effective cases **back-door speech acts** – speech acts that, so to speak, operate through the back door. While the explicitly uttered sentence is negotiated in the foreground, presuppositions can simultaneously establish evaluations, hierarchies or relations of authority without themselves becoming the subject of the debate.

Consider the sentence:

> "We must reduce the burden on laboratory animals as far as possible within the framework of economic utilisation."

What is explicitly asserted here is, at first sight, something quite positive: burdens are to be reduced. Yet the formulation may simultaneously carry along another assumption in the background: **economic utilisation constitutes the frame within which any weighing takes place at all.** This priority is not itself given any grounds. It therefore also need not be explicitly defended. Whoever merely discusses how the burden is to be minimised within this frame may be tacitly adopting the frame itself.

It is precisely here that David Lewis's idea of **accommodation**, upon which Langton draws, becomes important. Conversations function, among other things, because participants continually adjust their shared background to what is said. Certain presuppositions, if no one problematises them, are, as it were, taken up into the conversational record. Langton is interested in the fact that problematic norms or claims to authority can be stabilised along this path as well. They need not be adopted by vote. It can suffice that they are treated as self-evident in the further course of conversation.

This makes the discourse-pragmatic form of de-addressing harder to perceive than an open rejection. Were someone to say, "The suffering of these animals is of no consequence," the normative claim would be easy to identify and to attack. A conversational order in which no one utters this sentence, but in which deliberation is organised such that animal suffering may only appear under already fixed categories – such as degree of burden, threshold value, or proportionality – can be far more effective. The decisive question is then not refuted. It is **simply never admitted as a question capable of being decided** in the first place.

Langton also shows, however, that such pragmatic processes are not irreversible. Listeners can extract a presupposition being carried along and make it explicitly the subject of discussion. For this she draws on Marina Sbisà's concept of **explicitation**. What was casually presupposed becomes an open question: *Wait a moment – why does this assumption hold in the first place?* Langton's example is a sexist chant at a football match. A listener responds not by discussing the explicitly voiced call directed at the player, but by exposing the demeaning attitude toward women hidden within it: what, exactly, is supposed to be wrong with being a girl? In this way, a speech act that had previously exerted its effect unnoticed can lose its back-door character.

Langton calls this counter-operation **blocking**. For the topology developed here, a terminological distinction is important: **in Langton, blocking is precisely not de-addressing.** Blocking is the counter-speech that prevents a problematic presupposition from being tacitly accommodated. De-addressing, on our fourth layer, consists rather in a **preclusive discursive blockade** that prevents an objection from altering the conversational record – and thereby frequently prevents Langton's blocking from succeeding in the first place.

This also shows why counter-speech can achieve more than a refutation. According to Langton, blocking can, under certain circumstances, respond to a problematic speech act not merely with a contrary sentence, but by altering its **felicity conditions**. When a presupposed authority or background assumption is explicitly contested, the original speech act can pragmatically fail – in Austin's terminology: *misfire*. Counter-speech then does not merely say, "That is false." It can bring about: **"This sentence no longer functions as self-evidently as it was meant to."**

For debates in animal ethics and ecology, this distinction is especially important. For the counter-evidence here often does not originate from a being that can itself object at the discursive table. The animal can display panic, develop behavioural disorders, or resist an experimental arrangement; an ecosystem can degrade, species can vanish, and processes can become unstable. But it is human speakers who must first carry these occurrences into deliberation. The question then is not only **whether their statement is true**, but whether the discursive order permits this statement to alter the shared frame of the problem.

An idealised commission case makes the structure visible. A participant says:

> "These dogs, when kept in isolation, develop pronounced behavioural disorders and panic."

The chair responds:

> "Thank you for sharing your personal impression. We note that the well-being of the animals has been raised, and now turn to the standardised burden indices."

The statement was not disputed. No one showed that the observed behavioural disorders do not exist. Nevertheless, something decisive has occurred pragmatically. By being labelled a **"personal impression,"** the observation is placed on the side of the subjective; the subsequently mentioned **"standardised burden indices"** are, by contrast, granted the status of properly objective material. An epistemic hierarchy has arisen without ever having been argumentatively defended. The objection remains on the record – but it has lost its possible status as a **correction of the basis for decision**.

Miranda Fricker's reflections on hermeneutical injustice reveal a related danger. An utterance may appear unreasonable or lacking in credibility to a listener because its style of expression and social background do not correspond to the forms the listener has learned to recognise as rational. In her example, Marge Sherwood's emotional or intuitive expression appears less reasonable to Herbert Greenleaf, while another listener might well recognise in it a thoroughly comprehensible objection. Fricker's counter-ideal therefore demands reflexive attentiveness to **why something appears incomprehensible or unreasonable to me** before I infer, from this appearance, anything about the quality of what has been said.

In the context of animal ethics, this insight carries an additional point. Whoever speaks of an animal's "panic," "desperation," "grief" or "fear" can easily be classified as anthropomorphic, emotional or sentimental. Such criticism can be justified; anthropomorphic misinterpretations do exist. It becomes preclusive only when the classification **replaces the examination**. "Sentimental" then ceases to be the result of an analysis and becomes instead a barrier to entry: precisely the claim that might need to be investigated is removed from the domain of serious reasons by means of the label.

The same pattern can be modelled in environmental procedures. A large-scale project is examined in a procedure with regard to land use, species protection, noise, compensatory measures and other defined criteria. Such categories are necessary. A preclusive blockade would arise, however, if an objection to the destruction of a historically evolved ecological fabric could only be translated into the question of **how many compensation points are required**. The original question – whether precisely this fabric is replaceable, and what is lost with its loss – would thereby not be answered. It would have been cast into a form in which it can no longer fundamentally disturb the procedure.

The problem, then, is not simply that procedures draw boundaries. Every discourse requires rules of relevance. Even the sentence "This does not belong here" can be entirely justified. A topology of de-addressing must therefore not describe every delimitation of a topic as an abuse of power. What matters is a more demanding diagnosis: **can it be contested, with reasons, how the boundary was drawn?** Can an objection show that the very relevance rule in force excludes a substantively important consideration? Or does the procedure protect itself from correction by only admitting objections to its frame within that very frame?

Herein lies the preclusive structure. Unlike in Layer 3, a concept is not necessarily missing. The objection is formulable and comprehensible. Unlike in Layer 2, an empirical observation is not necessarily missing either. The suffering or the ecological loss may be well documented. The blockade arises **between utterance and effect**: what is said does not receive the discursive status it would need in order to alter the further deliberation.

The counter-operation therefore consists in **unblocking**. Background assumptions must be made explicit, rules of relevance must be kept accountable to justification, and the very boundary between "objective" and "not objective" must itself remain open to critique. This does not require accepting every objection. It merely requires that an objection be capable of failing on grounds – and not already fail because of a pragmatic device that decides in advance that it does not count as a reason at all.

Yet this repair, too, can fail. A **counter-blockade** that in turn morally disqualifies every other position, or that breaks off the conversation altogether, does not establish a better discursive order. Langton's blocking is interesting precisely because it does not necessarily require silencing the speaker. It can suffice to make what had hitherto been taken for granted visible, and to halt its automatic accommodation. Only then does the actual debate begin.

Layer 4 thus possesses a clear diagnostic criterion: **de-addressing is present wherever a relevant objection can enter the discourse but can no longer reach its presuppositions.** The objection is recorded without being permitted to correct anything. Re-addressing, by contrast, does not mean granting every objection its due. It means keeping a discourse open enough that what is being indirectly asserted through animals or ecological realities receives at least the chance **not merely to be heard, but to alter the question itself**.

## 06 | Translation & Chains of Representation: The Problem of the Loss of Difference

What happens to a concrete living being when its faring is translated, step by step, into findings, measurements, statistics, indices, and finally administrative decisions?

With the fifth layer, de-addressing reaches a point at which the previous levels may already have been successfully traversed. The animal was perceived as a living bearer of expression. The data-gathering situation was methodically open. Adequate concepts were available, and objections could be raised within discourse. And yet something decisive can still be lost at the end. For science, law, and administration cannot operate directly with the immediate experience of an individual animal or with the concrete multi-layeredness of a habitat. They must **translate, select, standardize, and condense**. It is precisely these necessary translation steps that constitute a new possible site of de-addressing.

The problem, therefore, is not simply **abstraction**. Without abstraction there would be neither science nor a capable administration. A veterinary finding must convert individual observations into diagnostic categories; an epidemiological study must aggregate many cases; environmental planning requires maps, indicators, and comparable magnitudes. The question, rather, is this: **What happens to the differences that cannot be carried along at each of these stages?** Does a path back to them remain open — or does the necessary compression gradually turn into a representation in which the original phenomenon appears only as an administrative quantity?

One may imagine such a chain of representation as a series of translations. An animal breathes heavily, moves differently, withdraws, or shows recurring behavioural disturbances. From this a veterinary finding arises. Many findings are subsequently aggregated into prevalences, mean values, and risk classes. These, in turn, feed into an operational welfare index, a certification system, or a political indicator. No single translation need be false. Herein lies precisely the difficulty: **a loss of difference can result from a sequence of operations that are each, taken on their own, factually reasonable.**

Bruno Latour’s concept of **articulation** offers a criterion by which this difficulty can be specified more precisely. In *How to Talk About the Body?*, Latour opposes the notion that scientific quality consists chiefly in depicting a reality as unchangedly as possible. A merely inarticulate representation, for him, at the limit merely repeats the same thing: “A is A.” A good scientific arrangement, by contrast, brings forth new relations and differences: “A is B, is C, is D.” It allows a phenomenon initially determined only roughly to become **more differentiated** through further instruments, concepts, and possibilities of comparison. Scientific progress, then, is shown not by our managing with ever fewer concepts, but by our learning to register ever finer differences.

Latour’s starting example is the training of a sense of smell. Before training, various odours may appear almost identical; through smelling kits, teachers, chemical contrasts, and practice, a body develops that perceives differences to which it was previously, so to speak, “mute.” *Learning to be affected* therefore means: **more reality becomes accessible in that more differences are enabled to become effective.** Instruments and artificial experimental arrangements are by no means enemies of authentic perception here. They can be precisely the means by which an initially impoverished world is articulated more richly.

Herein lies an important correction to an overly simple critique of quantification. A cortisol value, a movement sensor, or a computer-assisted behavioural model does not de-address an animal merely because it is technical or quantitative. A new sensor may reveal a difference hitherto overlooked. A statistical procedure may detect connections that would scarcely be discernible in the individual animal. Latour’s alternative is therefore not **experience instead of data**, but rather: *which new differences can be articulated by means of these data?* An instrument is good when, after its use, more can be distinguished than before — not when it merely reproduces the same original category with greater decimal precision.

In this connection, Latour speaks, with William James, of the **Multiverse**. This does not mean an arbitrary multitude of subjective “worlds.” The Multiverse is, rather, the world prior to its various articulations having been prematurely unified into a single privileged description. The alternative to premature unity thus does not consist in relativism, but in allowing different scientific approaches to remain effective alongside one another long enough that their differences are not already smoothed over in advance. Latour formulates the point concisely: the more we learn, the more differences can exist — more precisely, the more differences become registrable and effective for us.

But it is precisely here that it also becomes clear that **articulation is not yet the same as moral addressing**. A scientific system can be extraordinarily differentiated. It can continually take up new parameters, improve measurement procedures, integrate unexpected findings, and revise its models. And yet, within this ever richer landscape of data, something can still be lost: **the concrete being as bearer of its own faring**.

Imagine an animal-welfare research programme that develops ever better sensors. Movement profiles, body temperature, heart rate, vocalizations, feed intake, stress hormones, and social behaviour are continuously recorded. The model improves with each generation. Misclassifications lead to new parameters; new findings alter the weighting of the index. Epistemically, this would be a remarkably capable, learning system. In Peirce’s sense, the way of further inquiry would by no means be blocked. His famous research maxim states: **“Do not block the way of inquiry.”** For Peirce, it is precisely the readiness to relinquish existing convictions in the face of new experience that belongs to the will to learn. McLaughlin accordingly emphasizes that Peircean inquiry is iterative, self-correcting, and dependent on the capacity to replace old habits with new ones.

This is a strong epistemic norm — but not yet a sufficient theory of addressing. For a model could be entirely open to improving its **stress index** from version 4.2 to version 4.3 without ever asking what happens to a concrete animal whose suffering disappears within the acceptable average. The system then learns a great deal about stressors and can nonetheless increasingly see only **stress events**. The pig appears as a bearer of measured values, no longer readily as the individual living being *to whom* things are going badly.

This is precisely where the **residue of addressing** lies: after every legitimate abstraction, it must still be possible to ask **who or what is being represented by the representation**. In the case of the animal, the question runs: does an individual being remain reconstructible as the bearer of its own well-being and its own harm? In the case of the ecosystem, the question differs, since no experiencing subject may be presupposed here: does the concrete configuration, with its historical, spatial, and relational particularity, remain recognizable — or has it been fully dissolved into interchangeable functional quantities? This difference should be explicitly identified as a problem of the translation chain.

An example may make the process visible. Let the starting point be a single fattening pig showing restlessness, injuries, and recurrent tail-biting. At the first stage, we encounter a concrete animal in a concrete situation. At the second, selected aspects are technically recorded: movement patterns, lesions, feed intake, or other parameters. At the third stage, such data are combined with those of many animals and condensed into an operational welfare score. At the end stands, say, a figure such as **72 out of 100 points**. The score may be methodologically excellently constructed. Nevertheless, the statement “this animal is suffering here and now” has become a statement about the statistical quality of an operation.

The critical point is not that the later statement would be untrue. **The statements have different objects.** An operational index answers a different question than the observation of an individual animal. De-addressing arises only where this shift becomes invisible — that is, where “the operation achieves an acceptable value on the index” practically entails: “the animals here are kept acceptably.” The aggregate then begins not merely to summarize the individual findings but to represent them normatively. The individual animal may still be present as a data point, yet it no longer possesses any independent path back into the decision.

The translation functions similarly in ecological interventions. A species-rich, nutrient-poor grassland is, at first, a concrete place with particular soil, particular vegetation, populations, relations, and its own history of formation. For a planning procedure, this place must be mapped. Habitat types are determined, indicator species counted, and properties assessed. This information can then be converted into a points system. In the end, the loss can be expressed arithmetically as a certain number of compensation points and, where applicable, offset by measures taken at another location.

Here, too, it is not the mapping itself that is the problem. On the contrary: it may at first **articulate the habitat more richly**, since it renders visible differences that the casual glance does not perceive at all. The turning point occurs where the representation becomes a **logic of substitution**. Two habitats then count as equivalent insofar as their administrative values are mutually offsettable. What the index precisely fails to represent — site-specificity, historical formation, particular interactions, or irreversible losses — has scarcely any remaining possibility of correcting the equation. Arithmetically, everything is balanced; factually, something irretrievable may nonetheless have disappeared.

This allows the loss of difference characteristic of this layer to be specified more precisely. It does not simply mean that information is lost with every abstraction. That is unavoidable. The loss becomes de-addressing when three conditions coincide: the higher levels of representation become increasingly action-guiding; the differences excluded in the process can no longer be played back; and, finally, the compressed representation is treated as though it exhausted the morally or practically relevant object. The translation chain then becomes a one-way street.

Latour’s concept of articulation suggests a plausible countermovement to this. Good chains of representation must possess **return channels**. A statistical anomaly must be able to lead back to the individual case; a contradicting individual case must be permitted to disturb the rule of aggregation; an administrative compensation must remain open to review as to which concrete ecological differences it precisely does not replace. The aim is not to return from numbers to some supposedly unmediated “pure experience.” No such immediate level is available to institutions at all. The aim, rather, is a translation that knows its own losses and permits **counter-translations**.

This can be presented as a tension between Peirce and Latour. Peirce formulates a **minimal norm of openness**: an inquiry must not block its own further path of correction. Latour adds an **intensifying norm of articulation**: good research does not merely keep revision abstractly possible, but arranges its instruments so that ever new differences can become noticeable at all. The project of de-addressing demands, beyond this, a third question: **can these differences be traced back to the being or configuration for which they are relevant?**

Precisely for this reason, the counter-operation must not consist in a mere **increase in data**. More sensors, more variables, and larger data sets can be valuable. But they can equally produce an **inflation of articulation**: the representation becomes ever denser, while the reference to what is represented grows ever weaker. A barn with billions of data points per day is not, for that reason alone, better addressing than a careful observation of a few animals. A hundred ecological indicators do not guarantee that the loss of a concrete habitat remains adequately visible. “More articulation” in Latour’s sense therefore does not mean an arbitrary data overkill. It means **more relevant differences capable of bearing consequences for our description**.

The burden of proof on this layer therefore lies with the mediating instances

## 07 | Practice & Arrangement: The Materially and Spatially Constructed Exclusion

How can an animal disappear from our field of moral perception without anyone claiming that its wellbeing is meaningless?

At the sixth layer, de-addressing becomes **material**. Thus far the discussion has concerned perception, experimental arrangements, concepts, discourse rules, and translation chains. Now something moves into focus that easily appears as mere background: **spaces, paths, walls, fences, workflows, and material arrangements**. They determine who encounters whom, who can hear or see something, and which parts of a chain of action can even become perceptible at the same time.

De-addressing can therefore literally **be lodged in walls**.

This is philosophically significant because it exposes the inadequacy of a widespread assumption: that moral exclusion must always be locatable somewhere as a conviction—for instance, as a statement such as "animals do not count." A practice can rather be carried by people who by no means believe that animal suffering is irrelevant, and yet be organised such that this suffering **practically fails to appear** to most of those involved. The exclusion then lies not primarily in the minds of the agents, but in the architecture of their relations.

An everyday example is the meat counter. Between a living pig and a vacuum-sealed portion of pork lies a long chain of spatial and practical transformations: breeding, husbandry, transport, slaughter, cutting, refrigeration, packaging, logistics, and sale. At the end of this chain, the buyer no longer encounters an animal's body, but a standardised consumer product. Blood, voice, movement, gaze, smell, and the animal's spatial environment have vanished. No act of deception was required for this. **The infrastructure itself has dismantled the encounter.**

This does not mean that every division of labour or spatial separation is already morally problematic. Modern societies could hardly function without such differentiations. Hospitals separate operating theatres from public areas, laboratories require controlled zones, food production requires hygiene and logistics. The decisive question is therefore not: *Do separations exist?* It is: **which morally relevant connections are pulled apart so durably by the separation that they can no longer appear together in practice for those acting?**

### Visibility is spatially organised

The first vector of this de-addressing is **spatial segregation**. Animal husbandry, slaughter, and large parts of animal experimentation take place in spaces not intended for everyday public traffic. This can have good functional reasons: hygiene, disease control, occupational safety, or controlled experimental conditions. Yet these same precautions simultaneously carry an epistemic and moral side effect. They determine **who can come into bodily contact with which animal, and under what conditions, at all**.

A pig in a closed housing system, a mouse in a barrier breeding facility, or a cow in a slaughterhouse is, for most people, not simply difficult to assess. It is, first of all, **not present**. The everyday points of contact lie elsewhere: the meat counter, the medicine, the cosmetic product, the scientific publication. Between the living being and the user of its products lies a spatial infrastructure.

For precisely this reason, visibility must not be treated solely as an individual problem of attention. One can hardly reproach people for overlooking something when societal spaces are organised such that it normally does not confront them at all. De-addressing here is not a mistaken perception of a present counterpart. It is the **organisation of its absence**.

The counter-model can be seen with particular clarity in Sue Donaldson and Will Kymlicka. Central to their political theory of domesticated animals is that these animals are already members of shared societies. Membership shows itself not only in abstract rights, but in relations of dependency, cooperation, and shared use of space. Domesticated animals can form bonds with humans, coordinate activities, and share physical space—and it is precisely this proximity that, for Donaldson and Kymlicka, belongs among the preconditions of possible co-citizenship.

Space thereby acquires a political significance: **who is permitted to appear where?**

Donaldson and Kymlicka's discussion of dogs in public parks illustrates this particularly vividly. Conflicts over dogs are not only about faeces, noise, or safety risks. Behind them often stands the unspoken question of whether public space can be a shared human–animal space at all. The demand for dog-free public areas can rest on the notion that publicness is by its very nature a human space, and that animals belong in the private sphere of their keepers. The debate over dog parks thereby becomes a dispute over **who belongs to the shared world and may help shape it**.

This insight can be generalised beyond Donaldson and Kymlicka's own examples: spatial order does not merely depict social belonging. It **produces and stabilises** it.

#### No one performs the whole act

A second mechanism is **temporal and functional decoupling**. Modern production processes break activities down into specialised partial operations. In meat production, one person handles breeding, another fattening, transport, stunning, slaughter, cutting, inspection, packaging, and sale. Each person takes on only a segment.

Division of labour, too, is not inherently de-addressing. It enables expertise and can even improve accountability. It becomes problematic where the division of the action simultaneously effects a **division of its moral object**.

The veterinarian sees the health status of a herd. The transport company sees fitness for transport and logistics. The slaughterhouse sees carcasses. Quality control sees threshold values. The trade sees goods. The consumer sees a food product. Each segment may be correctly grasped in itself—and yet no one, within their own practical horizon, needs to trace the entire movement from the animal born to the packaged product.

Here layer 6 differs from the preceding translation layer. There, the question was how an animal is **represented** when behaviour becomes data and data becomes indices. Here, the question is how practice itself is distributed such that **no one any longer stands at a point from which the entire chain of action could be experienced**.

The individual animal thus disappears twice: epistemically in the representations, and practically in the organisation of labour.

This structure also complicates the attribution of responsibility. The question "Who does this?" increasingly receives the answer: no one alone. Everyone performs only a functionally limited part. This does not mean that no one is therefore responsible. But responsibility must then itself be **reconstructed across the division of labour**. A topology of de-addressing is particularly helpful here, because it does not search for the one guilty agent, but asks how a practice interrupts its own lines of connection.

#### When a body becomes a product

The third vector is **material-aesthetic transformation**. It becomes especially clear with foods of animal origin. In the course of processing, an animal's body changes not only its location but its mode of appearance.

The animal becomes a carcass, from which cuts are made, then portions, ingredients, and products. In a steak, its origin from a body can still be recognised relatively easily; in sausage, minced meat, or nuggets, this connection is considerably weaker. Packaging, refrigeration, standardisation, and product design remove further traces of the living organism.

This is, at first, a genuine process of transformation, not merely a linguistic veiling. Precisely for this reason it is of interest to the topology. At layer 3, a concept such as "livestock" could categorially narrow perception. Here something similar happens **materially**: the object itself is altered such that certain descriptions subsequently suggest themselves almost naturally. A neatly packaged portion of meat truly appears as merchandise, because the entire material prehistory in which it was still the body of a living being has been removed from its present form.

The commodity does not lie. But it **tells almost nothing more of its origin**.

This makes the distinction between deception and de-addressing important. No person is needed who actively conceals that meat comes from animals. The material transformation can itself ensure that this knowledge plays scarcely any role in the concrete moment of action. One *knows* that the cutlet comes from an animal, without standing before an animal when reaching into the refrigerated shelf.

Thus the concept of "experience without a voice" acquires a spatial-practical inflection: the animal's experience not only lacks discursive self-representation. The very being whose experience would need to be represented has been **physically removed** from the situation in which its product is disposed of.

#### The laboratory: a protected space and an epistemic filter

Laboratory spaces are similarly ambivalent. Protective clothing, airlocks, controlled air circulation, standardised cage systems, and regulated contact times serve genuine scientific and hygienic purposes. They must therefore not be hastily interpreted as mere instruments of exclusion.

At the same time, they shape the possible relationship to the experimental animal. Someone who encounters an animal only during fixed test periods, under standardised conditions, and within an instrumental protocol encounters it differently from someone who shares an ongoing everyday life with the same animal. Certain differences become scientifically more visible; others may recede.

Here layer 6 connects with Despret's problem of the observational situation: it is not the individual test alone, but already **the architecture of the laboratory**, that determines which relations are probable, desired, or excluded. The decisive question is therefore not whether laboratory spaces are "artificial." It is: *which forms of responsiveness does this space enable—and which does it render practically impossible?*

Donaldson and Kymlicka's reflections on animal agency offer an interesting contrast here. They emphasise that an animal's capacities for participation do not reside in isolation within the individual, but depend on relationships, opportunities, and shared spaces. Agency is supported or suppressed through "uptake," through recognition, and through suitable environments. Precisely for this reason, they call for providing spaces in which animals can genuinely explore their preferences and forms of cooperation under safe conditions.

The point can be sharpened for the topology: **a space does not merely house capacities. It can either draw them out or make them disappear.**

#### Fenced-in nature

A related question arises in environmental policy. Nature conservation inevitably works with spatial boundaries: protected areas, core zones, buffer zones, wildlife corridors, and zones of use. Such boundaries can be ecologically necessary. De-addressing therefore does not arise simply through a fence or through the designation of a protected area.

What becomes problematic, rather, is a **logic of segregation** according to which nature possesses a strong claim only within explicitly designated spaces, while the remaining space appears as a matter of course a human space for use. The establishment of a reserve can then, paradoxically, stabilise the view that outside this reserve, nature need only satisfy residual functional demands.

The spatial pattern reads:

**Here nature—there use.**

It is precisely at this point that a topology of de-addressing differs from a romantic demand for boundless wilderness. Boundaries can protect life. What matters is whether they **enable relationships or sever responsibilities**. One wildlife corridor connects separated habitats; another fence can prevent the very same movements. The physical structure is similar, its relational meaning opposite.

#### From separated spaces to shared spaces

## 08 | Institution & Politics: Co-Citizenship without a Voice and Political Representation

How can beings be politically taken into account who are affected by our decisions, but who cannot themselves formulate their claims in parliaments, courts, or hearings?

With the seventh layer, the topology reaches the level of **institutions**. The previous layer concerned spaces and material arrangements: Who can encounter whom, who is spatially separated, and which chains of action are practically pulled apart? Now the question is: **What happens when a morally relevant objection reaches a political or legal decision? Is there even a place there where it can become effective?**

With animals, this problem is particularly evident. Domesticated animals do not live outside human societies. Human beings have bred them, integrated them into their economic and forms of life, made them dependent, and regulated almost all conditions of their lives legally. Donaldson and Kymlicka therefore speak not merely of morally protection-worthy beings, but of **de facto members of our political communities**. Precisely domesticated animals find themselves physically within society and at the same time under human dominion. Their habitats, reproduction, freedom of movement, labour, medical treatment, and finally, often also their death, are determined by human institutions.

Here a peculiar asymmetry arises: animals are **objects of political decisions** without being able to be political speakers in the usual sense. They write no statements, elect no representatives, and conduct no administrative proceedings. It does not follow, however, that they can assert nothing. Their behaviour can reveal preferences; they can withdraw, cooperate, form attachments, show resistance, and react to changed social conditions. The political difficulty therefore does not lie in the fact that “nothing is said.” It lies in the fact that modern institutions are built above all for a very specific form of objection: for the **propositionally articulated claim of a speaking citizen**.

De-addressing at this layer is therefore typically **privative**. An institutional way back is missing. An animal may have become quite clearly visible at an earlier layer, its behaviour may be well documented scientifically, and its condition may be adequately described conceptually. Nevertheless, all of this can remain politically without consequence if no institution is obliged to take up this evidence, transform it into a legal claim, and, where appropriate, revise a decision. The blockage indicator here therefore reads: **There is no effective channel of revision.**

#### Must a citizen be able to argue?

Donaldson and Kymlicka's most important intervention consists, first, in relieving the concept of **citizenship** itself of a certain burden. A traditional picture of political membership presupposes a citizen who can reflect on rules, exchange reasons, and consciously commit to shared norms. In a particularly pointed form, this line of tradition extends as far as an opposition between the rational human being and “animal” lack of self-control.

Donaldson and Kymlicka reconstruct this picture partly by reference to Plato and Rousseau. In this tradition, political freedom possesses a special value precisely because rational human beings can master their drives. Animals, by contrast, embody the counter-image of unrestrained freedom. In Plato, dogs, horses, and donkeys populate the democratic city as an emblem of a freedom that has lost its limits; in Rousseau, political freedom is bound to the capacity not to be determined merely by impulse. The citizen thereby appears as someone who has **overcome the animal within himself**.

Donaldson and Kymlicka, however, consider this picture misleading both empirically and politically. The everyday functioning of a democracy by no means rests on citizens steering every encounter through conscious moral-philosophical reflection. The greater part of social cooperation functions through practised consideration, trust, habitual self-restraint, empathy, and an embodied sensitivity to social norms.

Human **agency** therefore also possesses a “bodily” and socially distributed side. It arises in relationships, responses, and mutual recognition. Donaldson and Kymlicka take up here the idea that agency does not reside solely in a sovereign will, but rests on *uptake* by others: an action also acquires its social meaning through the fact that others take it up and react to it.

This changes the threshold of political membership. Whoever demands that a citizen must be able to analyse and explicitly justify shared norms propositionally excludes not only animals. The same threshold would also affect small children, people with severe cognitive impairments, dementia, or temporarily limited rationality. Citizenship would thereby become fragile or even impossible for many human beings. Donaldson and Kymlicka regard this as a confusion: the point of citizenship is not to award a prize for special cognitive achievements. Citizenship rather designates **membership in a common, politically regulated society** whose public good must also take into account the subjective well-being of its members.

They therefore explicitly draw on developments in the theory of children's rights and disability studies. There too, political participation was increasingly detached from the idea that only fully autonomous, rationally arguing adults could politically assert their interests. Participation must rather be organised in such a way that different forms of expression and dependency are taken into account. Donaldson and Kymlicka see in this not a mere special case, but a step towards a **more inclusive theory of citizenship**.

The analogy should not, however, be overstretched. Animals are not children, and people with cognitive impairments are not animals. The philosophical point concerns solely the institutional threshold: **the absence of propositional self-representation must not be equated with the absence of interests, the absence of agency, or the absence of political membership.**

#### Animals as co-shapers of social norms

Donaldson and Kymlicka go a step further still. Animals are not merely to be **represented** better, as if they were passive wards. At least in the case of domesticated animals, they see forms of their own participation in shared social practices.

Their criticism is directed against two opposing images. According to one, animals are so pliable that human norms can be imposed on them almost arbitrarily. According to the other, they are “unruly beasts” — uncontrollable creatures who fundamentally cannot understand or help shape rules. Both notions, according to Donaldson and Kymlicka, share the same error: they conceive social norms as exclusively human constructions, to which animals have only the alternative of **obedience or disobedience**.

Against this, they propose the possibility of animal agency. Domesticated animals can not only learn to cooperate with human beings. They can, through their behaviour, help determine **which cooperation actually works**. Good social norms do not then arise exclusively from human beings determining how a dog, horse, or sheep is to behave. They must also respond to which relationships the animal seeks, refuses, or alters. Donaldson and Kymlicka therefore speak of creating conditions in which animals can safely try out different forms of cooperation.

The example of the dog in the park makes this idea especially vivid. A dog learns consideration: not to jump up at people, to avoid conflicts, to respect the boundaries of others. But human beings, too, must change their behaviour. They must, for example, allow opportunities for sniffing, movement, and social interaction. In the successful case, the rules of a shared space do not arise from the one-sided disciplining of the animal, but from **micro-adaptations on both sides**.

This gives rise to a precise concept of political agency without a parliamentary voice: the animal formulates no ordinance. But it can **generate evidentiary pressure**. Its behaviour makes visible that a rule works or fails, that an environment is bearable or burdensome, that a cooperation is accepted or refused. Political representation must institutionally ensure that this non-propositional objection can be translated into propositional and legally effective forms.

#### From object of protection to political point of reference

Here lies an important difference between **care** and **representation**. Whoever understands animals merely as objects of protection asks: *What should we do for them?* An addressing institution must additionally ask: **Which of their reactions may correct our decision?**

This is more demanding. For human representatives cannot simply claim to be “speaking for the animal.” The asymmetry does not disappear through representation. On the contrary: because animals cannot correct their representative in a debate, the latter's accountability increases. An animal advocacy office would therefore have to bring together various forms of evidence: observed behaviour, veterinary findings, ethological knowledge, individual life history, and — where possible — repeated preference tests. Its task would not be to put a fictitious human voice into the animal's mouth, but to **keep open the way back from the institutional judgement to the levels of expression and experience**.

With this, layer 7 connects to the preceding levels. The institution must not simply assert an “animal will.” It must be able to verify whether its representation still fits with bodily expression, the situation of data collection, the available concepts, and scientific translation. Political representation is thus the highest stage of a long chain of representation — and precisely for this reason particularly subject to revision.

#### Three political relations instead of a single animal status

Donaldson and Kymlicka further propose not to place animals into a uniform political category on the basis of a single moral status. What is also decisive are the **relations** in which different groups of animals stand to human societies.

For **domesticated animals**, they use the model of **citizenship**. Human beings have bred these animals into their societies, made them dependent, and integrated them into shared orders of cooperation. From this, claims to membership arise.

**Wild animals**, by contrast, stand fundamentally in a different relation. They are not to be integrated into human societies, but to possess claims to territorial autonomy. Donaldson and Kymlicka use for this the political category of **sovereignty**.

In between stand **liminal animals**: pigeons, foxes, rats, and other non-domesticated animals that live permanently in human-shaped spaces without being members of a cooperative commonwealth in the same sense. Donaldson and Kymlicka compare their position to **denizenship** — a status of residence without full citizenship.

Precisely this three-pillar model guards against a new form of de-addressing: political inclusion must not mean **including all beings according to the same model**. A wolf is not better addressed by being declared a citizen of a human commune. Its appropriate political relation may consist precisely in human institutions respecting its territorial independence.

This is also important for ecological contexts. Transferring Donaldson and Kymlicka's concept of citizenship to an **ecosystem as such** would be an additional theoretical step that their texts do not undertake. An ecosystem possesses no animal agency and no subjective well-being. The project can, however, transfer the underlying institutional question: **which legal mechanisms ensure that ecological changes can actually cause political decisions to be revised?** The analogy thus concerns the channel of revision, not the claim that ecosystems are citizens.

#### When participation is merely played out

This also brings into view the failure form of this layer: **ritualised participation**.

Institutions can create bodies of representation, conduct hearings, or appoint animal welfare officers, and thereby formally create the impression of political consideration. What is decisive, however, is whether this participation can have **consequences for decisions**.

## 09 | Theoretical Form & Philosophical Deflection: The Academic Evasive Maneuver

Can a philosophical theory de-address animals even though it explicitly speaks about animal ethics?

On the eighth and uppermost layer, the topology is directed against the **form of thinking itself**. Until now, de-addressing could be localised at relatively concrete points: in a perception, an experimental arrangement, a concept, a discourse rule, a translation chain, a spatial arrangement, or a political institution. Now the more difficult question is posed: whether the **basic architecture of a theory** already decides what can count at all as morally visible, as an admissible reason, or as philosophically relevant experience.

This is a different kind of blockade. A theory need not forbid any observation and need not explicitly reject any objection. It suffices if its basic concepts determine in advance the form an experience must take in order to count philosophically. If, for instance, only propositional reasons, quantifiable preferences, or certain cognitive capacities are accepted as access to moral relevance, a filter arises **prior to** any particular argument. What does not fit this format then appears not as a refuted objection, but as something that has not yet been properly articulated philosophically at all.

De-addressing on this layer is therefore **constitutive**: it is not a faulty application that damages an otherwise neutral theory. The blind spot can already lie in the very way the theory generates its object.

#### When reality overwhelms thought

Cora Diamond starts from an experience that initially resists translation into the language of philosophical problems. With her formula of the **“difficulty of reality”** she designates situations in which something actual seems to elude our habitual forms of comprehension. This is not simply a particularly complicated problem for whose solution we still lack information. Rather, Diamond describes experiences in which the mind gets the sense of being unable to encompass something — a reality that does not refute our habitual concepts but, as it were, **knocks them off their hinges**.

Her point of departure is a poem about six young men who appear fully alive in a photograph and were killed shortly afterwards in the First World War. Of course the state of affairs can be explained without contradiction: the photograph was taken while they were alive; later they died. And yet precisely the simultaneity of overwhelming vitality and certain death can produce an experience that does not disappear through this correct explanation. The difficulty then lies not in a logical contradiction, but in **what it means truly to let this reality in**.

For Diamond, our relationship to animals also belongs to this domain. In her reading of J. M. Coetzee's *The Lives of Animals*, Elizabeth Costello does not simply represent a philosophical position to be defended with arguments. Costello is affected, bodily and spiritually, by what human beings do to animals. Diamond speaks of a “profound disturbance of the soul” and takes seriously that Costello's relation to the exploitation of animals is presented as a wounding of her own body. The difficulty consists precisely in continuing to live with this knowledge in a world in which the killing of millions is simultaneously a social norm.

This is something other than the question:

**“Which moral theory provides the best argument against meat consumption?”**

The question can be legitimate. It can even be important. But it is not identical with the experience to which Coetzee's text draws attention.

#### Deflection: from horror to soluble problem

At this point Diamond adopts from Stanley Cavell the concept of **deflection** — the evasive movement or diversion. Deflection occurs when a difficulty of reality is translated into a philosophical problem **in its vicinity**, and one then proceeds as though this had captured the original difficulty.

Diamond sees precisely this in philosophical reactions to Coetzee. Costello's disturbing experience becomes a familiar debate in animal ethics: May we eat animals? Do animals have rights? Is the analogy Costello uses admissible? Which arguments are coherent? Which statements are based merely on emotion? Diamond puts it pointedly: philosophy knows how to handle such questions. Universities teach precisely how to examine difficult arguments, to justify claims, and to distinguish emotional distortions from rational reasons.

The problem is not that these would be poor philosophical abilities. Diamond's actual point is more subtle. **The difficulty of argumentation can replace the difficulty of reality.**

One then works in a highly demanding way — but on a different problem.

For Diamond, the difference is visible precisely in the body. Philosophical deflection can transform Costello's wound into a further fact: she is emotionally affected; her being affected can now itself be examined as to whether, and to what extent, it is morally relevant. Thereby the body, which was previously the site of the disturbance, becomes a further datum within the theoretical analysis. Diamond describes the alternative as an “inhabiting of a body”: not registering the fact of a wounded body, but attempting to grasp the difficulty at all from an embodied perspective.

For reflections on **experience without a voice**, this is a central point. Theory does not de-address only when it explicitly devalues animals. It can also de-address by speaking very reasonably about animals while thereby altering the very form of what was originally supposed to be responded to.

#### “Anything but Argument?” — What counts at all as moral thinking?

Diamond had already examined this problem earlier in *Anything but Argument?*. Its starting point is the claim that a serious moral appeal must ultimately function through **arguments**. Anyone who wants to convince people to treat animals differently must not merely address their feelings, but must provide reasons.

Diamond does not dispute the value of arguments. She opposes the stronger assumption that only argumentatively reconstructible paths of persuasion can constitute **rational moral thinking**. Literature can change our moral perception without its content being fully translatable into a sequence of explicit premises and conclusions.

Her example of Dickens shows what is at stake. Dickens does not simply convey additional facts about children. He brings readers to perceive a world **from a child's perspective**. Diamond describes the particular form of attention thereby cultivated: an affective, imaginative interest that lets us recognise a human being as the centre of their own view of the world. The moral significance does not lie only in information that could subsequently be fed into a theory. It already lies in the **manner of seeing** that the description enables.

Correspondingly, a cold focus on "the facts" can be extraordinarily rich in information and yet represent a morally impoverished form of attention. Diamond contrasts this with a warm, concentrated, and humorous attentiveness to the lives of others. In successful literature, such description contributes to our lasting sense of what is interesting and important in human life.

Applied to animals, this does not mean replacing science with empathy. The point is rather: **even the capacity to recognise what scientifically established facts mean at all possesses a dimension of perception.**

A cortisol level does not by itself say what fear is. A lesion does not by itself explain what it means that *this animal* is injured. And even a fully correct description of a slaughter process still does not guarantee that we perceive the event under a morally adequate description.

#### Alice Crary: whoever neutralises the world must add morality back in later

Alice Crary systematises a related thought in *Inside Ethics*. Her diagnosis is directed against a metaphysical presupposition widespread in modern moral philosophy: the idea that the objective world is, as such, **morally neutral**. Moral qualities, on this view, do not belong to empirically given reality. They must be added by valuations, preferences, practical reason, or moral principles.

Crary calls the corresponding metaphysical presuppositions **“hard metaphysics.”** Whoever accepts them has already methodologically determined that an empirical description of humans and animals must initially manage without moral qualities. Science supplies us with the facts about bodies, behaviour, brains, or preferences; ethics then begins with the question of what moral status these facts ground.

It is precisely this architecture that Crary describes as placing humans and animals **“outside ethics.”** Not because the respective theory would necessarily be hostile to animals. On the contrary: a theory can ground very strong animal rights. What matters is the order:

**First the value-neutral being — afterwards the moral qualification.**

Crary's counter-position consists in seeing humans and animals **“inside ethics.”** Her strong thesis is that they possess moral properties that are observable in an empirical sense. Moral perception is thereby not understood as a subjective addition to an otherwise finished objective world, but as a possible way of grasping this world itself more fully.

This does not mean that every moral judgement is thereby automatically true. Moral perception, too, can be biased, sentimental, superficial, or simply wrong. Crary does not defend an infallibility of moral sensibility. Her point concerns rather the **conception of objectivity**: there is no reason to exclude a description from the realm of objective knowledge simply because its adequate understanding requires moral perspectives. In her further argument, she therefore defends a "wider" conception of objectivity, within which ethically informed modes of thought can make an internal contribution to real understanding.

This has methodological consequences. If moral qualities are not in principle excluded from the empirical world, literature, precise description, ethology, history, and other forms of attentive perception can jointly work on determining **what kind of life we actually have before us**. Crary explicitly criticises the widespread philosophical practice of outsourcing empirical knowledge entirely to supposedly ethically neutral disciplines. In her view, literature and the humanities can themselves contribute to a world-guided understanding, precisely because they open up different moral perspectives.

#### Moral status or moral visibility?

Here the particular function of the eighth layer can be grasped more precisely. Many animal ethics begin with the question of **moral status**:

*Which property must a being possess in order to require moral consideration?*

Sentience? Consciousness? Interests? Autonomy? Future-oriented preferences?

These questions can be important. However, the topology draws attention to the fact that even their placement can already contain a prior decision. For they possibly presuppose that a morally neutral being is initially present, and that a property must subsequently be sought which **qualifies it into** the moral space.

Crary's critique reverses the direction of view: perhaps we do not first have to prove that this animal counts morally before its fear, attachment, vulnerability, or dying may appear as morally significant. Perhaps a living being, under suitable descriptions, already meets us in such a way that the question of its welfare possesses moral significance **within** our perception.

This is not a finished animal ethics. It does not yet decide what we ought to do.

But it shifts the burden of proof.

#### Utilitarianism: inclusion and possible deflection

Precisely for this reason the critique of utilitarianism must be nuanced. It would be wrong to treat utilitarian animal ethics simply as de-addressing. Historically, it was Peter Singer in particular who generated enormous visibility for animal ethics precisely by asserting the sentience of animals against anthropocentric exclusions. Utilitarianism can thus perform a powerful re-addressing operation

## 10 | Extension to Ecosystems and Animal Species: Systemic Limits of Resonance

Can de-addressing be extended beyond sentient individual animals to populations, species, and ecosystems – even though these can neither suffer nor speak nor raise claims of their own?

With this extension, the object of the topology changes fundamentally. In the case of the individual animal, the starting situation was comparatively clear: **there is experience, but no discursive self-representation.** An animal can experience pain, fear, well-being, attachment or unease, and can indicate through behaviour and corporeality what is happening to it. The ethical task consists in translating this experience, in a controlled manner, into human processes of cognition, justification and decision-making.

In the case of an **animal species**, a **population**, or an **ecosystem**, matters stand differently. A species does not feel pain. An ecosystem has no inner perspective. A bog does not wish to be preserved, and a river does not fear its own pollution. If one wished simply to extend the model of animal experience further and further, one would have to ascribe to such entities a quasi-subjectivity that they do not possess.

Precisely for this reason, the extension to environmental ethics is not a mere prolongation of the previous scale. It requires a **change in the mode of addressing**.

Here one should distinguish between **indirect addressing** and **mediated addressing**. In the case of the animal, the moral point of reference is already present: a being with a welfare of its own. It merely cannot represent itself in the space of reasons. In the case of species, ecosystems, landscapes or rivers, by contrast, it must first be determined **what is even to count as an addressable unit and in what its vulnerability consists**. They become morally visible through process structure, historical genesis, ecological function, dependencies, significance for the future, and specific forms of loss. The "empty chair" of an experience without a chair is not here merely occupied vicariously; it must first be methodically constructed.

This is not a gradation of moral dignity. The three forms of direct, indirect, and mediated addressing designate different **problems of translation**, not different quantities of morality. It is precisely this distinction that prevents a naive gradient thesis, according to which the less a being is able to speak, the more strongly human beings would have to speak on its behalf, until eventually stones, mountains and climate systems would be endowed with an unlimited duty of representation. That would be an overextension of the concept of addressing. The relevant question is not: **how mute is something?** But rather: **what real structure, vulnerability, or form of loss makes this something morally relevant in the respective constellation?**

#### From the animal to the population to the species

Already the transition from the individual animal to the population shows why several levels must be kept apart.

A wolf population consists of sentient individuals. Each individual wolf can experience hunger, pain, or fear. In this respect, the classical perspective of animal ethics remains intact. At the same time, the population can possess properties that belong to no individual animal: population size, genetic diversity, age structure, reproductive rate, spatial distribution, or long-term viability.

A measure can therefore be burdensome for individual animals and favourable for the population – or the reverse.

Thus two morally relevant levels stand side by side. The population must not be elevated to a "whole" so strongly that the suffering of individual animals disappears. But neither may the individual animals determine the view so completely that long-term population dynamics become invisible. Already at this point, an ethics of individual sentience alone no longer suffices without further ado.

In the case of a **species**, the situation shifts once again. A species is not a large animal. Its extinction is not the death of a supra-individual subject. For this reason, a purely pathocentric language applies here only indirectly. This limit should be made visible: if moral status is bound exclusively to individual sentience, species, populations, biotopes and ecosystems fall outside the immediate focus, or receive at most a derivative significance.

But this does not yet show that species extinction would be morally meaningless. It shows, for the time being, only that **a different description of value is required**.

A species can be understood as a historically evolved form of life: as a specific way of finding nourishment, reproducing, using space, being connected with other life forms, and responding to environmental conditions. When a species disappears, therefore, it is not merely an entry that vanishes from a taxonomic list. A particular, historically emerged possibility of living world-formation comes to an end. At this point, biodiversity should accordingly be described not as the greatest possible number of species, but as **plural, grown, and not arbitrarily interchangeable difference**.

#### Ecosystems: loss without a suffering victim

The shift in ethical logic becomes even clearer in the case of the ecosystem.

An ecosystem can be damaged without the ecosystem itself feeling anything. A river can collapse ecologically without suffering. A bog can be destroyed without formulating a claim. The language of experience, interest, and suffering must therefore not simply be carried over metaphorically.

It is precisely here that the problem of **"loss without a victim"** begins.

When a historically grown patch of poor grassland disappears, a bog is drained, or a complex river system is converted into a functionally monotonous channel, something real can be lost even though there exists no overarching subject to whom this loss occurs as an experience. Such environmental entities cannot therefore be characterised as hidden persons, but rather as **relational and processual configurations**. Their loss can consist in the fact that connections break apart, possibilities of life disappear, or a configuration loses its capacity to sustain diverse forms of life and to productively absorb future change.

This should be connected with the distinction between **continuants** and **occurrents**. Ecological objects frequently possess both aspects: populations, species, or biotopes can be traced over time as persisting units; at the same time, ecosystems consist of events, rhythms, disturbances, migrations, successions, and transformations. Whoever conceives of an ecosystem merely as a stable thing misses its processual character. Whoever, conversely, treats it merely as a sequence of changing events loses the possibility of speaking at all of the persistence or destruction of a particular ecological configuration.

Environmental addressing must therefore hold on to both: **identity without rigidity, and change without arbitrariness.**

#### Luhmann: nature does not communicate with society

At this point, Niklas Luhmann provides an entirely different, yet for this project particularly incisive, diagnosis.

Luhmann does not first ask what moral value ecosystems have. His question is: **how can a functionally differentiated society respond to ecological changes at all?**

The starting point is radical: a dying forest, a poisoned river, or a collapsing population does not communicate with society. Physical and biological events remain, initially, **the environment of societal communication**. For them to become socially effective, they must be translated into communication.

And this translation never occurs "for society as a whole." For Luhmann, modern societies consist of distinct functional systems – science, politics, law, the economy, and others – each of which can process ecological disturbances only according to its **own modes of operation and codes**. The differentiation of these systems expands their capacities, but at the same time limits what they can become sensitive to at all. Luhmann calls this their **capacity for resonance**.

"Resonance" here therefore means neither empathy nor moral attention. It designates the possibility that an event outside a societal subsystem **triggers connective operations of its own within that system**.

An ecological change can be biologically enormous and remain almost without consequence for society – as long as it finds no connection within the relevant forms of communication.

#### Natural frequencies: why not everything can become relevant everywhere

Luhmann uses a particularly apt metaphor for this: functional systems are capable of resonance only within certain **natural frequencies**.

The point can be illustrated by an ear. Precisely because a sense organ does not react to everything, it can be extraordinarily sensitive within a particular range of frequencies. Selectivity is therefore not merely a deficiency. It is a precondition of the capacity for perception.

Something similar holds true at the societal level. Functional systems must reduce complexity in order for anything to be processed as information at all. Luhmann therefore explicitly emphasises both sides: coding excludes an immediate, point-for-point relevance of the environment; but it is precisely this same reduction that first makes it possible for selected environmental changes to appear within the system as information.

This protects the topology from an overly simple critique of societal abstraction. It would be absurd to demand that the law must immediately reflect every ecological difference in juridical terms, that the economy must directly register every biological change, and that politics must instantly transform every scientific disturbance into a decision.

A system that reacted to everything could react specifically to nothing.

De-addressing therefore does not already begin with **selectivity**. It begins where selective societal processing leads to a relevant ecological change **finding no sufficient path into decision-capable communication**.

#### Science can recognise without deciding

The scientific system illustrates this limit particularly well.

Luhmann determines its code as **true/untrue**. Science can investigate ecological relationships with ever greater precision, develop models, correct measurements, and discard previous explanations. In this sense, it possesses a high capacity to translate environmental changes into scientific communication.

But it is precisely this specialisation that sets a limit. From the truth of a statement, no decision yet follows, for the scientific system, as to **what ought to be done**. Luhmann explicitly emphasises that the code of truth does not serve to enforce or select actions. This task must be taken up in other societal contexts.

Thus a society can know a great deal about the loss of biodiversity and yet respond politically hardly at all.

This is not merely an inconsistency on the part of individual persons. Between

**"We know that something is happening"**

and

**"We are therefore changing our practice"**

lie several system translations.

It is precisely here that Luhmann connects with the preceding layer of **representation chains**. Even more perfect data guarantee no moral or political addressing. Scientific resonance is necessary, but not sufficient.

#### Politics: first address – but no ecological central steering

Luhmann's analysis becomes particularly interesting with respect to politics.

Ecological problems are quickly **addressed politically**, because political systems can produce collectively binding decisions. Whenever something is to be changed socially, politics therefore appears as the obvious address.

Luhmann warns, however, against deriving from this a general societal competence of politics for everything. Politics, too, possesses its own operational logic. It cannot act outside its own conditions. The political system, too, is therefore "capable of resonance only within the framework of its natural frequencies."

Moreover, political decisions do not have immediate ecological effects. A parliamentary decision is, initially, communication, and generates further societal operations – such as laws, administrative measures, or financial incentives. Only through long chains of effect does a habitat perhaps change.

## 11 | The Peircean Meta-Norm: The Unblocked Path of Address

What prevents a scientific description, a welfare index, a legal regulation, or a political institution from eventually becoming so self-evident that what it no longer captures can no longer return at all?

At the end of the topology, there is no further layer. Charles Sanders Peirce instead supplies a **counter-vector** that runs transversally through all the previous levels. His famous rule states:

**"Do not block the way of inquiry."**

Peirce formulates it in connection with his first rule of reason. Whoever wishes to learn must actually **want to learn** — and must therefore not be satisfied with what he is already inclined to believe in any case. For Peirce, the real danger to rational thought does not lie in trying out a false hypothesis. Errors can be corrected. More dangerous is a manner of thinking that itself blocks the path of future correction.

For the topology of de-addressing, this is a surprisingly powerful meta-norm. At every previous level, selection, abstraction, translation, and simplification were unavoidable. No one can perceive an animal's entire behaviour at once. No experiment can capture every possible variable. Language operates with categories; administrations need indicators; political procedures require jurisdictions; theories must distinguish what they consider relevant. **Selection is unavoidable.**

De-addressing therefore does not begin wherever something is excluded from view.

It begins where what has been excluded **can no longer return**.

A provisional abstraction becomes problematic when it declares itself the final description of reality. An administrative indicator becomes de-addressing when no individual case can any longer disturb the indicator. A theory becomes dogmatic when counterexamples are indeed registered but are, in principle, prevented from altering its foundational categories. And a political institution closes off the path of addressing when new evidence can indeed arrive but has no institutional way back to a decision already made.

The Peircean imperative gathers all these cases under a single common question:

**Where can what we have excluded raise its objection again?**

#### Four ways of blocking the path

Peirce describes four particularly characteristic ways in which philosophical thought blocks its own capacity to learn. His examples are not drawn from animal or environmental ethics. Precisely for this reason, they are of interest as a meta-diagnosis.

The first blockage is the **absolute assertion**. Scientific convictions are treated as though they already possessed final certainty. Peirce's fallibilism is not directed against settled judgements; research must provide provisional answers. The conviction becomes problematic where "According to our best current knowledge, this is so" imperceptibly turns into: **"It cannot possibly be otherwise."**

In animal research, this blockage might appear, for instance, in the claim that a particular behaviour has been conclusively explained as a mere reflex. In environmental policy, an established assessment procedure might be treated as the definitively adequate description of ecological quality. In each case, the critical point does not lie in the fact that a position is held. It lies in the fact that new observations are no longer permitted to be the kind of thing by which the position could, in principle, be changed.

For Peirce, the second blockage consists in claiming that something can **in principle never be known**. His famous example is Auguste Comte, who declared the chemical composition of the stars to be fundamentally unknowable — shortly before spectral analysis opened up precisely this path of knowledge. Peirce's point is directed against the leap from present ignorance to knowability in principle.

For "Experience Without a Voice," this warning is especially important. There are, of course, epistemic limits to understanding animal experience. We cannot simply know "what it feels like" for the animal. But it does not follow from this that refined ethology, behavioural observation, physiology, and relational research could reveal nothing new about it. The claim **"We can never know this about animals anyway"** can therefore itself become a form of epistemic closure. It transforms methodological modesty into a prohibition on knowledge.

The third blocking strategy declares something to be an **irreducible ultimate fact**. Peirce criticises conceptions according to which certain basic elements of an explanation are simply fundamental and therefore not open to further investigation. To call something "inexplicable" is, he argues, not yet itself an explanation.

In political and institutional contexts, the same figure of thought frequently appears under the name of **necessity** ("Sachzwang"). Economic necessity, technical inevitability, or institutional unavoidability may, of course, refer to genuinely real constraints. They become Peirceanly problematic where the constraint loses its own prehistory: *Why exactly are these costs unavoidable? Which alternatives were excluded? Which institutional order produces this necessity in the first place?* Where such a question is cut off in principle, the alleged necessity functions like Peirce's philosophical ultimate fact.

The fourth blockage concerns, finally, the claim that a law, a regularity, or a theory has reached its **final and perfect formulation**. Peirce recalls Laplace's famous verdict that stones do not fall from the sky — although meteorites had, of course, fallen. Unusual phenomena are then not treated as a challenge to a theory but are declared impossible by virtue of the theory itself.

Here Peirce directly touches upon the eighth layer of the topology. An anomaly is interesting precisely because it opens up the possibility of revising not merely a single answer but the very **order of questioning** itself. If unusual animal behaviour, new ecological developments, or unexpected damage may only be classified as disturbances of a fundamentally correct model, the way back is blocked.

#### The way back is not an additional layer

This is why Peirce does not belong at the top of the vertical architecture as a ninth or eleventh station.

The Peircean way back runs **through** it.

At the level of **body and expression**, it means: an established interpretation must allow itself to be disturbed by new behaviour.

At the level of the **research situation**, it means: an animal must retain the possibility of not reacting the way the experiment expects it to.

At the level of **conceptual stock**, it means: categories must remain capable of being supplemented and de-absolutised.

In **discourse pragmatics**, it means: presuppositions and relevance rules must themselves be capable of being made explicit and contestable.

In the **chain of translation**, it means: there must be a path leading back from aggregates and indices to those differences that were lost in the process of aggregation.

At the level of **practice and arrangement**, it means: spatial divisions and divisions of labour must not render the overall connection permanently unrecognisable.

At the level of **institution and politics**, it means: new evidence must actually be able to reopen a decision.

And at the level of **theoretical form**, it means, finally: reality must be able to alter even the categories with which we attempt to grasp it philosophically.

The way back is therefore not a path back to some supposedly unmediated origin. We cannot simply strip away scientific, conceptual, and institutional mediations and arrive once more at the "pure animal" or "pure nature." What is meant is something different: **every mediation must remain correctable by that which it mediates.**

#### Peirce and Latour: keeping open and becoming more differentiated

At this point, the relationship between Peirce and Bruno Latour also becomes clear.

Peirce first formulates a **minimal norm**. He demands, negatively, that no legitimate path of inquiry be finally blocked. The existing description must remain, in principle, correctable.

Latour's concept of articulation is more demanding. Good research should not merely remain open to eventually being refuted. It should arrange its instruments in such a way that **more differences can become visible and effective at all**. His example of the trained sense of smell shows: the gain in knowledge consists in becoming sensitive to contrasts that could not previously be distinguished. A good research process does not render its object ever more uniform, but increasingly articulated.

One might therefore say:

**Peirce protects the way back. Latour demands that something genuinely new can actually come back along this way.**

Peirce: *Do not close off inquiry.*

Latour: *Learn to register more differences.*

The two norms complement one another. Mere openness can remain empty: a procedure can be theoretically correctable at all times without actively seeking out differences that might disturb its categories. Conversely, Latour's increase in articulation requires a Peircean minimal condition: if the system is closed against revision, new articulations can no longer change anything.

McLaughlin's reconstruction of Peirce's first rule underscores precisely this connection between fallibilism and learning. For Peirce, inquiry is an iterative, self-correcting process; the "will to learn" begins with dissatisfaction toward the present state of one's own knowledge.

#### But revisability is not morally sufficient

Precisely at this point, however, the topology must **go beyond** Peirce.

For "Do not block the way of inquiry" is, in the first instance, an epistemological norm. It does not yet say **toward whom or what we ought to be morally attentive**.

A system of animal husbandry could be exceptionally Peircean in its organisation. It could continually test new sensors, correct its statistical models, take in unexpected data, and discard established hypotheses. An algorithm could learn from every new observation. The system would be epistemically open.

And yet it could still treat the animal exclusively as a **bearer of quantified production events and stress events**.

The path of research would be open, but the path of **addressing** would not thereby necessarily be open as well.

This is one of the most important conceptual distinctions of the entire model:

**Revisability is a necessary condition of addressing — but not its content.**

Peirce prevents dogmatic closure. He does not yet say what must be preserved within an open procedure.

For animal ethics, this **residue of addressing** reads as follows: the concrete animal must remain reconstructible as a being with a welfare of its own. Its burden must not be entirely dissolved into the fact that a model correctly processes its data.

For ecological systems, the residue is different. Here there is no suffering supra-individual subject. What must be preserved, instead, is the possibility of asserting real ecological differences — historical formation, relationships, functional interconnections, irreversible losses, or diminished generative stability — once again against a social representation.

Peircean openness is thus the **form** in which correction remains possible. What is to count as a morally relevant correction must be determined additionally by the respective ethics.

#### From meta-norm to practical test procedure

From this connection, a simple test procedure can be derived for research and institutions that remain open to addressing.

The first question concerns the **basic layers of expression and observation**:

**Can the being under investigation still surprise us?**

In Despret and Stengers, this is called recalcitrance