Animal ethics should consistently accord animals direct moral status.

Why should we not torment animals? An obvious answer is: because the animals suffer as a result. But one can also argue differently: cruelty to animals is to be rejected because people who treat animals cruelly become brutalised themselves. Or because the animal belongs to someone. Or because an animal species is important for an ecosystem.

All of these justifications can lead to animals being protected. But they do so for different reasons. It is precisely at this difference that the first metaethical norm is directed.

What does "directly morally relevant" mean?

Something is morally relevant if it may not simply be disregarded in our moral deliberation. Direct moral relevance further means: direct moral relevance means that the animal counts for a reason that lies within itself.

This can be tested by means of a simple thought experiment. Let us imagine that the additional human reason were to fall away. Would the animal still count?

A dog, for example, is not protected merely because it is a human's property. Even a stray dog may not be tormented. Its pain does not vanish from the moral landscape simply because no one asserts a claim of ownership over it. If its capacity to suffer itself provides a reason to take it into consideration, then its moral relevance is direct in this sense.

The matter is different in the case of a justification that runs exclusively through another bearer of value. In that case, the animal is indeed protected, but indirectly: the actual reason lies elsewhere.

Kant: protection of the animal for a human reason

A classic example is provided by Immanuel Kant. For Kant, human beings have duties not to treat animals cruelly. The animal is thus by no means left to human arbitrariness.

The decisive reason, however, does not lie in the animal itself. Cruelty towards animals can dull human compassion and thereby damage moral conduct towards other human beings. The duty to treat animals gently is therefore, for Kant, an indirect duty: the animal is protected because something concerning human morality is at stake in how we deal with it.

The first metaethical norm regards this construction as insufficient. For one may ask:

What if the cruel treatment of an animal did not cause a human being to become brutalised?

If the moral reason against animal cruelty falls away along with the risk of human brutalisation, then this reason does not lie in the suffering animal itself. This is precisely what makes the justification indirect.

The norm does not thereby claim that Kant provided no moral protection for animals whatsoever. It asks, more precisely, where this protection is grounded.

Direct does not mean equivalent

An important distinction is needed here. Direct moral relevance does not imply that animals and human beings must be morally equivalent.

A consideration can be morally relevant without being decisive in every situation. If an animal suffers, this suffering can therefore constitute a genuine moral reason without thereby already determining how strongly this reason weighs against all other reasons.

The first norm therefore does not demand a tierethical egalitarianism. It does not require that: animals must count just as much as human beings.

It requires, initially, only this: animals must be able to count on account of what they themselves are and experience.

Whether different moral addressees are equivalent, and how competing reasons are to be weighed against one another, are further questions.

Why "consistently direct"?

An animal ethics may initially take animals into direct consideration and later lose this directness again. This is why the norm speaks not merely of "direct", but of consistently direct.

A theory must be examined at several points: How does it ascribe value to animals? What duties arise from this? Can animals possess their own rights-positions? And what happens when animal and human concerns collide?

A theory might, for instance, declare that an animal's suffering is in itself morally relevant. If, however, it subsequently gives human interests categorical precedence in cases of conflict without further justification, a question arises: is the directness of the animal's relevance, initially acknowledged, thereby withdrawn again at the level of weighing?

"Consistently" therefore turns the norm into a coherence test. It is not enough to describe animals as morally significant at some point within a theory. The theory must show what becomes of this acknowledgement throughout its entire justificatory structure.

A difficult case: compassion

It is precisely in the case of compassion that the concept of direct moral relevance proves itself to be in need of clarification.

When I see that an animal is suffering, I perhaps feel compassion. Why, then, does the animal count morally?

Two different answers can be given to this.

On a first reading, the connection is direct because the animal's suffering evokes my compassion. My response answers to something that befalls the animal. The animal would then be directly morally relevant: compassion is a reflex response to its experience.

One may, however, ask more strictly where the justifying reason lies. If the argument ultimately runs: "The animal's suffering causes my suffering, and therefore it should be ended", the decisive value seems once again to lie in my own state. Tom Regan raises just such an objection against an ethics of compassion: it is not my compassion that is to confer value upon the animal in the first place; the value must be anchored in the affected being itself.

This gives rise to two criteria for "direct": understood responsively, an animal ethics is direct if its moral response is triggered by a property or experience of the animal.

Understood in terms of justification, it is direct only if the reason for moral consideration lies in the animal itself, and is not mediated first through the state of a human observer.

The first reading is broader. The second is more demanding.

The metaethical task therefore does not consist in concealing this difference. Rather, an animal ethics should make clear which understanding of direct moral relevance it employs.

A linguistic trap: "indirect" does not always mean the same thing

In animal ethics, the word indirect has yet another meaning.

Animals can suffer, feel fear, experience wellbeing, or wish to avoid something. But they cannot themselves introduce these experiences into a philosophical discourse. Their experience must be perceived, described, and translated into reasons by human beings. This can be termed indirect address.

This indirect address is something different from an indirect justification of moral relevance.

An animal can thus be addressed indirectly while at the same time being directly morally relevant.

"Indirect address" describes the difficulty of translating animal experience into the space of human reasons. "Indirect moral relevance", by contrast, describes a theory in which the reason for protecting the animal lies with another bearer of value.

The two meanings must not be confused with one another.

What does the first norm require?

The norm does not yet answer the question of how we ought concretely to treat animals. It does not, for instance, say whether animal experimentation is always forbidden, whether human beings may eat meat, or what rights animals ought to possess.

It poses a prior question directed at the architecture of an animal ethics:

Can the animal count morally in this theory for a reason of its own – and does this reason remain intact when the theory proceeds to duties, rights, and conflicts?

The norm thereby safeguards an elementary insight: an animal must not become morally visible within an ethical theory only by virtue of being important for something else.

Test question

Anyone examining a tierethical position may therefore ask:

Would the moral reason for taking this animal into consideration persist if the mediating human reason were to fall away?

If it persists, this speaks in favour of direct moral relevance. If it vanishes, one should examine whether the theory grasps the animal only indirectly, in moral terms.

Animal ethics should examine justificatory adequacy and orientational adequacy separately.

A philosophical theory can be convincingly grounded and yet lead to moral demands that convince us little. Conversely, we may consider a certain form of dealing with animals morally appropriate even though the philosophical theory used to justify it has weaknesses.

Two questions must therefore be kept apart: How well is an animal ethics grounded? And: How convincing is the animal morality to which it gives orientation?

The first question concerns justificatory adequacy, the second orientational adequacy.

Animal ethics and animal morality are not the same

By animal morality we mean our moral judgements and orientations in dealing with animals: May one kill animals? Which forms of animal husbandry are defensible? When are animal experiments permissible? What do we owe a pet?

Animal ethics operates at a higher level. It examines how such judgements are grounded. Which properties of an animal are morally relevant? Which principles are employed? How do duties, rights, or prohibitions arise from them?

This distinction is more than a matter of terminology. For a theory and the morality it supports can fare differently well.

A seductive shortcut

According to Norm 1, an animal ethics appears especially convincing when it grasps animals directly in moral terms. The reason why the animal counts then lies with the animal itself.

One might easily be tempted to continue this thought: The better an animal ethics captures animals theoretically, the better must also be the animal morality that arises from it.

It is precisely this inference that Norm 2 blocks.

The quality of an ethical justification does not automatically entail the quality of the practical orientation associated with it. The two levels are connected, but they do not coincide.

Kant and Wolf: a cross-position

The difference can be illustrated by comparing two animal ethics.

In Immanuel Kant, animals are grasped only indirectly in moral terms. Cruelty toward animals is to be rejected, among other reasons, because it can blunt human compassion and morally coarsen human beings. According to Norm 1, this is problematic: the decisive moral reason does not lie with the animal itself.

Kant's animal morality can nevertheless, in many situations, lie astonishingly close to familiar moral judgements. Animals should not be tormented; their burden should be limited; a cruel treatment of them is condemned. A theoretically indirect justification can thus lead to an animal morality that appears plausible to us in parts.

In Ursula Wolf, the line runs differently. Animal suffering becomes immediately significant for morality. This means her ethics initially fares better with respect to the direct moral grasp of animals. Yet it is precisely the generalisation of compassion that can generate very far-reaching demands. Where the principle is universalised, lifeworldly differentiations come under justificatory pressure.

This gives rise to a cross-position: an animal ethics can be problematic in terms of justification yet morally plausible.

Another can be plausible in terms of justification yet appear problematic in its moral consequences.

This does not mean that Kant is therefore right and Wolf wrong. The comparison shows something methodological: an animal ethics must be examined on two different levels.

What does "intuitively convincing" mean?

Caution is needed here. That an animal morality corresponds to our intuitions does not prove that it is correct.

Moral intuitions arise within biographies and cultures. Human beings grow into forms of the human–animal relationship. That we treat dogs differently from pigs, pets differently from wild animals, or human beings differently from other animals appears self-evident to many. But self-evidence is not yet justification.

Common sense, that is, our pre-philosophically familiar moral convictions, can be mistaken. Cultures have, throughout their history, taken practices for granted that were later rejected for good reasons.

The reverse inference, however, would be equally premature: because moral intuitions are fallible, they are not meaningless.

When a philosophical theory regularly produces results that collide with well-considered moral experiences, an irritation arises. Such an irritation does not refute the theory. But it does create a new task of clarification.

One should therefore say neither:

"This feels right, so it is right."

nor:

"The theory is correctly grounded, so any opposing moral experiences are irrelevant."

Both would close the relationship between ethics and morality too quickly.

Moral orientation is work toward coherence

One way of understanding this relationship is coherentism.

Here, moral judgements are neither simply deduced from a supreme principle nor generalised from individual intuitions. Rather, theory, moral experience, concepts, knowledge, and concrete situations are repeatedly brought into relation with one another.

The glossary of my Applied Ethics (Introduction) refers to this work as coherence work.

Coherence does not mean that everything must ultimately fit together harmoniously. A theory can correct familiar intuitions for good reasons. Conversely, a persistent moral irritation can show that a theory has overlooked an aspect or abstracted too far.

What matters is that neither side wins simply because it is called "theory" or "intuition".

An ethics cannot simply produce morality

There is a deeper reason for this caution. Philosophical ethics supply reasons. But good reasons do not automatically entail a corresponding motivation. Human beings do not simply adopt their moral attitudes because they are presented with a valid argument.

Nor do we have fully transparent access to why certain values convince us, which motives actually drive our actions, or how moral attitudes come to change.

There is therefore no methodically secured coupling between the quality of an ethical justification and the motivational force of the morality developed from it.

A philosophically elegant animal ethics need not convince human beings. Conversely, a philosophically weak justification can support a moral practice that appears sustainable in many situations.

For this reason too, a metaethics must judge both achievements separately.

Does this threaten moral conservatism?

Norm 2 could be misunderstood. If our existing animal morality is recognised as an independent standard of examination, one might infer from this that customary morality ought to be preserved as far as possible.

That would be wrong.

Existing morality is neither given by nature nor unchangeable. The human–animal relationship has arisen historically and culturally. Precisely for this reason it can be criticised and changed.

Norm 2 does not protect the moral status quo, then. Rather, it protects the independence of moral experience from philosophical theory — and equally the independence of philosophical critique from common sense.

Theory can criticise morality. Moral experience can irritate theory. Neither side alone holds a monopoly on decision.

What does the second norm require?

Norm 2 does not require that an animal ethics fit our existing convictions as closely as possible. Nor does it require that philosophical theories weaken their demands as soon as these become inconvenient.

It requires a methodical separation: How convincing is the justification? How sustainable is the moral orientation arising from it?

Only once both questions are posed can one discern where a theory is strong, where an animal morality is strong, and where a tension between the two remains.

Test question

Anyone examining an animal-ethical position may therefore ask:

Is it being inferred prematurely here that, because the ethical justification is convincing, the animal morality following from it must also be convincing — or conversely, that because an animal morality is plausible, its philosophical justification must be sound?

If either of these occurs, justificatory adequacy and orientational adequacy are being confused with one another.

Animal ethics theories should not introduce animal-ethical egalitarianism without reflection.

Humans and animals have much in common. Humans can feel pain – so can many animals. Humans have needs and interests – so do animals. Some animals can plan, remember, form social relationships, or mourn their fellow creatures.

Such similarities are important for animal ethics. But does it follow from them that humans and animals are of equal worth?

This is precisely where the third metaethical norm comes in. It is not directed against animal-ethical egalitarianism. Perhaps humans and certain animals are indeed of equal worth. The norm merely requires that this equal worth not already be presupposed while it is ostensibly still being established.

What is animal-ethical egalitarianism?

Egalitarianism here means, first of all: human and animal individuals can be of equal moral worth.

That is stronger than the claim that animals are morally relevant. Norm 1 required that animals be grasped directly in moral terms. This means: animals should be able to count for a reason that lies in themselves. But it does not yet follow from this that they count equally much or in the same way as humans.

We must therefore keep three claims apart:

• An animal can be directly morally relevant.
• An animal can possess a different moral standing than a human.
• And an animal can be morally equal in worth to a human.

None of these claims follows readily from any of the others. The third norm thus keeps open a possibility that is easily lost in animal-ethical debates: directly relevant, but not necessarily of equal worth.

Equal worth is not equal treatment

Even among humans, equality does not mean that everyone must always be treated identically.

A simple example is people with disabilities. A person who cannot walk may need an elevator, special access, or additional support. They are treated differently from a person who does not require such aids.

This differential treatment does not contradict equal worth. It can be required precisely because of equal worth.

This shows something fundamental: equal worth does not demand uniformity.

Human beings can differ greatly in their abilities, needs, life circumstances, and possibilities and still be regarded as of equal worth. Equal worth is not derived from first establishing sufficient similarities between them.

On the contrary: because equal worth already holds, we then ask which similarities and differences are relevant in a given situation.

The direction of the justification

This changes the direction of the argument. One might initially argue as follows:

1. Humans can suffer.
2. Animals can suffer as well.
3. Humans and animals are therefore similar in a morally important respect.
4. Therefore, humans and animals are of equal worth.

Yet the step from 3 to 4 is not self-evident.

That two beings share a relevant property initially shows only a similarity. It must additionally be explained why equal moral worth follows from this similarity.

The equality argument in the case of humans often works precisely the other way around:

Because humans are regarded as of equal worth, we examine which of their similarities and differences may be relevant to a specific question.

In the case of the right to vote, for instance, sex is not supposed to make a difference. In the case of accessibility, by contrast, a disability can indeed constitute a relevant difference. In both cases, the presupposed equal worth is what first orients the question of which empirical differences count morally.

The direction of the argument therefore does not simply run

from similarity to equal worth,

but rather, frequently,

from equal worth to the moral significance of similarities and differences.

The charge of speciesism

This difficulty becomes especially visible with the concept of speciesism.

The expression is formed by analogy with racism and sexism. The underlying thought runs roughly as follows: if it is morally wrong to treat humans worse because of their skin color or sex, why should it then be permissible to treat animals differently solely because of their species membership?

This is a powerful moral challenge. It makes visible that the boundary between human and animal can itself stand in need of justification.

Metaethically, however, something further must be examined.

The charge of racism or sexism already operates within an order in which the equal worth of the humans concerned is presupposed. Precisely for this reason, one can ask whether skin color or sex may be relevant at all to a particular instance of differential treatment.

In the case of the human–animal boundary, however, this very equal worth is contested.

Whoever claims: "Speciesism is like racism" cannot therefore prove, by analogy alone, that humans and animals are of equal worth. For the analogy already operates with an assumption of equality whose extension to animals is precisely what is supposed to be established.

The danger is that the conclusion is unnoticedly already contained in the premises. In logic, such an error is called a petitio principii: one presupposes what one actually set out to show.

A criterion of relevance is not yet a criterion of equality

The difficulty becomes especially clear when a property such as capacity for suffering is used.

Let us assume: capacity for suffering grounds moral relevance. We then establish empirically: humans can suffer. Pigs can suffer. Dogs can suffer. Many other animals can suffer.

Something important follows from this at first: if capacity for suffering is morally relevant, the suffering of these animals may not simply be disregarded.

But it does not yet follow that: all beings capable of suffering are of equal worth in every moral respect. The relevance test answers the question of who or what must be taken into consideration. Egalitarianism answers a further question: what relation of worth exists among those who are taken into consideration?

Whoever conflates the two tacitly turns a criterion of relevance into a criterion of equality.

Singer: How far does the analogy carry?

Peter Singer's critique of speciesism can be examined precisely at this point.

Singer draws attention to the fact that humans and many animals possess interests and, in particular, want to avoid suffering. Species membership alone then does not seem to be a good reason to give comparable interests different consideration.

The strength of the argument lies in the fact that it generates a burden of justification: whoever weighs animal suffering less heavily must explain why.

But here too a metaethical question remains open. From the presence of comparable interests alone it does not follow that humans and animals are therefore bearers of equal moral status.

Interests can ground moral relevance. Whether they also ground equal worth is a further question.

The third norm requires that this additional step be made visible.

Regan: Equality can also be hidden within the criterion itself

Animal-ethical egalitarianism can enter a theory in yet another way.

Tom Regan describes certain humans and animals as subjects-of-a-life. Whoever meets this criterion possesses inherent value. The criterion functions like a threshold: one either belongs or one does not.

Such a construction can generate equal worth merely through its form. If everyone who crosses the threshold receives the same categorial status, an egalitarian relation arises among them.

This can be philosophically well founded. But here too Norm 3 applies: the egalitarian consequence of the status criterion must be made explicit.

A theory should not give the impression that it has merely discovered a property of animals, when the architecture of its criterion simultaneously produces a particular order of equality.

An overlooked possibility

Between two extreme positions lies an important third possibility. The first states: only humans are directly morally relevant. The second: humans and animals are directly morally relevant and of equal worth.

But it is also possible that: humans and animals are directly morally relevant without thereby being of equal worth.

This possibility is decisive for the connection between Norm 1 and Norm 3. Norm 1 prevents animals from being taken into consideration merely because humans have interests in them.

Norm 3 prevents equal worth from being made out of this direct consideration without additional justification.

This opens up broader moral terrain. One need not choose between the alternatives: either of equal worth or morally secondary.

Something can count morally for its own reason and yet count in a different way than something else.

Does this mean that animal-ethical egalitarianism is false?

No. The third norm is deliberately weaker than a refutation of egalitarianism.

One could, for instance, argue that certain great apes should be included, across species boundaries, in a conception of humanity and are therefore of equal worth to humans. In that case, equal worth would not simply be derived from individual similarities such as intelligence or capacity for suffering. Rather, it would need to be established why our conception of who belongs to this community of equals should be extended.

Such an egalitarianism, too, remains a possible position within animal ethics.

Norm 3 requires only: show whence the equal worth derives.

It must not vanish as an unnoticed additional assumption within an argument.

Two meanings of "egalitarian"

The expression egalitarianism itself also calls for attention. This section concerns the equal worth of those morally taken into consideration: are humans and animals of equal worth?

One can, however, also apply "egalitarian" to the structure of a theory. In that case, the expression means, for example, that various values or principles stand alongside one another and none rules over the others from the outset.

The two are not the same.

A theory can regard humans and animals as unequal in worth and nevertheless weigh several moral values against one another on an equal footing. And a theory can posit humans and animals as of equal worth while ordering its moral principles hierarchically.

Unreflective animal-ethical individualism blinds itself to important moral problems.

When an animal is in pain, it is always a single animal that suffers. A species does not suffer. An ecosystem does not feel fear. Nor can biodiversity be harmed in the way a sentient being can be harmed.

This seems, at first sight, to speak in favour of a simple animal-ethical consequence: what is morally relevant are individual animals, because only individuals can have experiences.

But it is precisely here that the fourth metaethical norm begins. It does not ask whether individual animals are morally important. That is not in dispute at all. Rather, it asks:

What disappears from an animal ethics when it can grasp only individuals as morally relevant bearers of value?

What is animal-ethical individualism?

An animal ethics is individualistic if its moral judgements are ultimately referred to individual beings.

This is, at first, very plausible. Pain, fear, stress or well-being are experiences of individual animals. A pig suffers, not "the species pig". A bird is injured, not a population as such.

Precisely pathocentric animal ethics — that is, ethical theories that place suffering at their centre — therefore have a strong individualistic tendency.

The connection is almost inevitable:

Suffering is an experience of individuals. Whoever ties moral relevance essentially to sentience thereby directs their gaze towards individuals.

This is not thereby already false. But it can result in other morally significant objects no longer becoming visible at all.

A theoretical decision that does not even look like a decision

The problem consists in the fact that an animal ethics frequently does not explicitly choose individualism at all.

It decides, for instance, in favour of sentience as the criterion of moral relevance. From this it follows almost automatically that only beings capable of suffering come into view.

• Species cannot suffer.
• Populations cannot suffer.
• Ecosystems cannot suffer.
• Biodiversity cannot suffer.

Thus these objects disappear from the moral field of view, without any decision having been made as to whether they might nevertheless be morally relevant.

Individualism is then not justified, but rather supplied along with the choice of the relevance criterion.

This is precisely what the norm means by "unreflective".

The blind spot

A theory has a blind spot if, owing to its own concepts and criteria, it can no longer even recognise certain moral problems as problems.

This is something different from a false answer.

If a theory says, for example:

"The preservation of an animal species is morally less important than the suffering of individual animals,"

then it has at least seen both considerations and weighed them against one another.

But if, owing to its concepts, it can grasp only individual suffering as morally relevant, then the value of an animal species does not even arise as a possible consideration in the first place.

Then no weighing has taken place.

Part of the moral terrain has already vanished from the map before the weighing has even begun.

The glossary to my Applied Ethics (Introduction) refers to such a structure as de-addressing. A possible moral addressee is not explicitly excluded. Rather, it does not appear at all as something that could be morally taken into consideration.

A touchstone: invasive animals in Australia

The problem becomes particularly clear in dealing with invasive animal species.

In Australia, introduced animal species threaten native populations and species. When human beings kill invasive animals, this can cause considerable suffering to the animals affected. At the same time, it is meant to prevent other populations from being pushed back or entire species from being wiped out.

A strictly individualistic ethics of suffering sees very clearly, at first: here sentient animals are being injured and killed. This is morally relevant.

But on the other side stand considerations of a different kind: the survival of endemic species, biodiversity, ecological relationships, or culturally significant landscapes.

These considerations are not themselves capable of suffering.

It is precisely for this reason that the case is metaethically interesting. It compels us to ask whether a theory fails to recognise moral problems solely because its vocabulary cannot express them.

A theory in which only individual suffering possesses moral value could, in fact, already have decided the conflict before any weighing has even begun.

Individualism is therefore not false

Norm 4 does not claim that animal-ethical individualism is false.

Individual animals are evidently central moral addressees of animal ethics. An ethics that protects biodiversity while ignoring the concrete suffering of individual animals would itself have a considerable blind spot.

The norm demands something more modest: an animal ethics should know what its individualism makes visible — and what it can render invisible.

Individualism becomes problematic when it turns from a perspective into a complete theory of what is morally relevant.

From the individual to the moral terrain

The Australian case therefore does not present a simple duel between two values. There is not merely: animal suffering versus species conservation.

The moral terrain can be far richer than that. Within it, individual suffering, biodiversity, ecosystems, scientific knowledge, cultural forms of life, and human responsibilities can all become relevant simultaneously.

A single theory may perceive some of these aspects with particular acuity. But precisely this strength can become a weakness if the theory mistakes its own particular section for the entire terrain.

The glossary describes such an error as pernicious abstraction: a single consideration is detached from a complex situation and then treated as though it exhausted the situation entirely.

Abstracting suffering is not, in itself, wrong.

The abstraction becomes problematic only when "suffering is morally relevant" unnoticeably turns into "only suffering can be morally relevant."

Individuals, species and ecosystems are not the same thing

The fourth norm thus also demands an ontological caution.

• An individual animal is something different from a population.
• A population is something different from a biological species.
• A species is something different from an ecosystem.
• And an ecosystem is something different from the individual organisms that live within it.

Whoever wishes to take all these objects into moral consideration must not level out their differences.

It would therefore be equally problematic to treat species or ecosystems simply as though they were large individuals. An ecosystem does not suffer, has no interests, and does not possess experience in the manner of an animal.

The alternative to unreflective individualism cannot therefore consist in declaring everything a moral subject.

Metaethical reflection must instead ask in what different ways different entities can become morally relevant.

Direct relevance need not always take the same form

This allows a connection to be drawn to Norm 1.

There, the demand was to grasp animals consistently as directly morally relevant.

Norm 4 now broadens this view. Even if individual animals are directly morally relevant, it does not follow that only individuals can be directly morally relevant.

• Perhaps an animal possesses moral relevance on account of its experience.
• A population might be relevant for different reasons.
• Biodiversity might, in turn, possess value in yet another way.
• An ecosystem might be taken into consideration on account of its structure, history, or generative capacity.

The differences between these forms must not be levelled out. Precisely for this reason, animal ethics should not too hastily seek a single characteristic that must be common to all morally relevant objects.

Pluralism rather than a new exclusivity

The alternative to unreflective individualism is therefore not necessarily collectivism.

A strong collectivism might merely reverse the error. The individual would then suddenly be subordinated to the preservation of the species, the ecosystem, or the whole. The concrete suffering of an animal might disappear, because only populations and systems still count.

The metaethical task therefore does not consist in replacing individualism with another exclusive perspective.

What is more plausible, to begin with, is a pluralism of sources of value:

• Individual animals can be morally relevant.
• Species and populations can also become morally relevant in certain constellations.
• Ecological relationships can constitute an independent consideration.

Which of these relevancies proves decisive in a concrete conflict is not thereby already determined.

It is precisely this open question that leads to the fifth metaethical norm.

Social structures, too, can escape one's view

Individualism harbours a second possible blind spot.

Whoever describes moral problems exclusively as relations between individual agents and individual animals may overlook the fact that our dealings with animals are organised within social structures.

A pig does not become a "farm animal" solely because a single human being has decided to treat it as such. Agriculture, the market, law, food culture, consumption habits, and institutions form relations within which animals are assigned particular roles.

The concept of the character mask draws attention to this level. Human beings and animals encounter one another within culturally evolved roles: as keeper and pet, farmer and livestock, researcher and laboratory animal, consumer and foodstuff.

Animal ethics theories should provide principles that make complex weighings possible.

If animal suffering is morally relevant, what follows from this?

Perhaps an animal experiment should not be carried out because it burdens animals considerably. Perhaps the same experiment nonetheless appears defensible if it offers the only realistic possibility of researching a serious illness. Perhaps the judgement changes again as soon as an animal-free alternative becomes available.

Such cases show that moral relevance alone does not yet determine what ought to be done.

This is precisely where the fifth metaethical norm comes in. An animal ethics must not only be able to explain why something morally counts. It must also be able to explain how different moral considerations can be related to one another.

Relevant does not yet mean decisive

The most important distinction is this:

Moral relevance is not the same as moral decisiveness.

A consideration is morally relevant if it may not simply be passed over in a moral deliberation.

It is decisive if it carries the overall judgement in a particular situation.

Let us take the suffering of a laboratory animal. If animal suffering is directly morally relevant, it may not be treated as though it were merely a technical drawback of the experiment. It belongs in the assessment as a moral reason.

This does not yet settle, however, whether the experiment must be prohibited.

Animal suffering can be relevant and nevertheless yield to other reasons in a particular constellation. Conversely, it can become decisive in a different constellation.

This distinction guards against two opposite errors. The first runs:

“Animal suffering is morally relevant, so animal suffering always decides.”

Then an important moral consideration is turned into an ultimate principle. The second runs:

“Animal suffering does not decide here, so it plays no moral role at all.”

Then non-decisiveness is confused with irrelevance. An adequate animal ethics must find a path between both errors.

The dimensionlessness of the moral point

An ethics can lose this capacity if its principles recognise only a single form of validity:

A duty is a duty. Full stop.

Then a principle either applies – or it does not. What is morally relevant appears at the same time as unconditionally required or forbidden.

Such a morality has, so to speak, lost its dimensions. It can indeed state very clearly what applies, but is increasingly less able to express how strongly a reason carries weight, with which other reasons it collides, and why it may be set aside in a particular situation.

The problem does not concern only a particular ethical school. Both deontological and consequentialist theories can formulate their principles in such a way that they ultimately function like decision-making machines.

Animal ethics, however, often needs more precisely this.

Human beings keep animals, eat animals, conduct research with animals, protect wild animals, kill invasive animals to protect other species, and use animals for medical purposes. In such situations, different considerations of value and duty collide.

An animal ethics that has already settled these conflicts through the very form of its supreme principle can indeed eliminate the complexity. But that may be precisely its problem.

Two paths – and two difficulties

One could try to extend the classical human-centred morality to include the concerns of animals. A first strategy runs:

We use several principles.

Rational persons, for example, possess moral status on the basis of their autonomy; sentient animals are additionally taken into consideration on the basis of their capacity for suffering.

This makes more morally visible. But now a new problem arises: what happens when both principles point in different directions in a concrete situation?

More principles do not yet produce a weighing. One must also be able to explain how they relate to one another. A second strategy runs:

We seek a more comprehensive principle.

Instead of reason, for example, capacity for suffering could become the decisive moral point of reference. Humans and animals would then fall under a shared principle.

This solves the first problem – but perhaps too well.

If the comprehensive principle already contains everything morally decisive, then in the end there is nothing left to weigh.

Precisely a theory that seeks to establish moral unity can therefore lose the capacity to differentiate. The source material diagnoses this in Ursula Wolf’s ethics of generalised compassion: if the principle applies universally – everywhere, always, for everyone and unconditionally – it becomes difficult to explain why, for instance, vegetarianism is demanded while certain animal experiments may nonetheless be permitted. Moral differentiation does take place, but the theory may no longer possess sufficient means to justify it itself.

Weighing does not mean arbitrariness

At this point there lies an obvious misunderstanding. If moral principles can be weighed against one another, one might think:

Then in the end anything can be justified.

An economic benefit, a tradition, or some human interest would only need to be weighted heavily enough, and animal suffering could then be set aside.

That would not be a sophisticated weighing, but interest opportunism.

Weighing means precisely that not just any reason may be traded off against any other. Reasons must be identified as reasons, and their significance depends on the concrete constellation.

From the sentence

“This animal experiment has scientific benefit”

it does not yet follow, for example, that:

“This animal experiment is justified.”

One must ask further:

• How significant is the expected gain in knowledge?
• How certain is it?
• How severe is the burden on the animals?
• Are there alternatives?
• Is the research question important enough?
• Can the number of animals be reduced?
• How does the situation change once new research methods become available?

Weighing is therefore not a relaxation of moral demands. It is a more demanding form of justification.

Prima facie and pro tanto: a reason remains a reason

This can be made more precise with two concepts from ethics: prima facie and pro tanto.

A prima facie reason is a moral consideration that must initially be seriously taken into account. “Prima facie” does not here mean “merely apparent”. The reason is genuinely present.

A pro tanto reason holds to the extent that its normative force carries. It can possess weight without alone determining the overall judgement.

Animal suffering is a good example of this.

If an experiment causes an animal pain, this speaks morally against the experiment. This reason does not disappear simply because the experiment may in the end nonetheless be permitted.

This is precisely what matters: an outweighed reason is not a refuted reason.

A justified animal experiment therefore does not become a morally neutral act. The suffering caused remains a counter-reason that must be retained in the overall judgement.

The language of prima facie and pro tanto thus prevents non-decisiveness from tipping back into irrelevance.

Reasons can point in different directions

C. D. Broad uses a further helpful notion for this: actions can possess right-making and wrong-making tendencies.

This refers to properties or aspects of an action that point in different normative directions.

One consideration can speak in favour of performing an action. Another can speak against it.

In the case of an animal experiment, for instance, the well-founded prospect of reducing severe human suffering might point towards permissibility. The suffering of the laboratory animals points in the opposite direction.

These reasons are not votes in a ballot, nor are they figures on a scale. The talk of “tendencies” is rather meant to express that:

Moral reasons have a direction, without alone already constituting the overall judgement.

The judgement arises from their combination in the concrete constellation. The glossary therefore speaks of directed normative vectors, from whose composition a judgement emerges.

A complex weighing therefore does not simply ask:

Which value is greater?

It asks:

Which considerations speak here for what, how far do they carry, and which judgement fits their concrete constellation?

Bridges between value and obligation

This brings yet another distinction into play. A duty does not follow directly from a value.

From

“Animal suffering is bad”

it does not yet follow without further ado that:

“Every action that causes animal suffering is forbidden.”

And from

“Biodiversity is valuable”

it likewise does not follow that:

“Every measure to preserve a species is permitted.”

Between a Concept of Value – a value judgement – and a Concept of Obligation – a judgement about requirement, prohibition or permission – lie additional argumentative steps.

The glossary to my Applied Ethics (Introduction) designates the connecting links required for this as bridge principles.

• One such bridge principle can be avoidability: if considerable animal suffering can be avoided without a comparably considerable loss, the moral reason to avoid it grows stronger.
• Another is reasonableness: a moral demand must take into account which possibilities for action and which burdens exist in a concrete situation.