The Orthodox View and Theories of Rights

All versions of the Orthodox View need to be augmented by some account of the notions of right and duty. Both are thorny concepts, but the notion of rights in particular has been vigorously contested. Especially in the English-speaking world, a typical starting point when theorising about rights is the analysis introduced by Wesley Newcomb Hohfeld.Footnote ²⁵ Hohfeld noted that the term “right” and other cognate terms are used in various, subtly different ways. For instance, it is normal to think that rights and duties are closely connected. However, there seem to be situations where we talk about rights without such a connection. We can say that Jane has the right to look at her neighbour’s garden from her house. Regardless, the neighbour is under no duty to allow this; he can for instance block Jane’s view by installing a tall fence. Hence, when we talk about Jane’s right here, we mean that Jane is permitted to view her neighbour’s garden. Hohfeld called such permissions privileges, and they are today usually called liberties or liberty-rights. On the other hand, claim-rights are rights that are correlated by duties. They are usually understood as the most central case of rights, and some scholars think that the word “right” should be reserved exclusively for claim-rights. However, the nature of claim-rights is not obvious.

Two main theories seek to explain what it means to hold a (claim-)right: the will theory and the interest theory. According to the will theory, holding a right means being empowered to make legally enforceable choices over the duties of others.Footnote ²⁶ For instance, if Maria owes some money to Ali, the latter has a right in this situation because he can for instance demand that Maria pay the debt, or choose to waive it. It is his choice. On the other hand, according to the interest theory, one holds a right if one’s interests are served by the duty of someone. Under the interest theory, Ali holds a right in the scenario as well, but for a different reason: because receiving money is, typically, in his interests (i.e. good for him). These theories come apart when we consider more controversial topics, such as animal rights. Many will theorists would claim that nonhuman animals cannot hold rights because they cannot, for instance, demand their rights, whereas animal rights are easier to explain under the interest theory: animals – at least sentient animals – certainly have interests, such as the interest not to suffer pain.Footnote ²⁷

The import of these theories for the discussion over legal personhood is not as straightforward as one might think. I will mention three issues here. First, one may ask whether any Hohfeldian legal position (such as a liberty) is enough to make its holder count as a legal person on the Orthodox View, or whether only claim-rights count.Footnote ²⁸ Second, one might think that the theories of rights also limit who or what can be a legal person: for instance, if one adheres to the will theory, one will plausibly deny that animals can hold rights or be legal persons. Hence – and assuming that animals cannot bear duties, either – animals cannot be legal persons. However, many people writing on legal personhood would, in fact, deny this conclusion. Instead, many subscribe to what I term the “Anything-Goes View”: legal personhood can be given to (virtually) anything. I will address this view in Section 2. Second, the Capacity-for-Rights View can easily be misunderstood. Philosophers or rights theorists might understand the “capacity for rights” as the capacity to hold rights in general.Footnote ²⁹ Hence, for instance, adult human beings of sound mind would be a paradigmatic example of someone with a capacity for rights, whereas nonhuman animals would not have the capacity for rights under the will theory. Such capacity does not depend on legal decisions; elsewhere, I have denoted it the “conceptual capacity for rights”.Footnote ³⁰ However, this is not how the “capacity for rights” is typically understood in the legal personhood context. Rather, the capacity for rights is understood as a legal status, whereby the legal system treats someone as a potential right-holder.

The Orthodox Inventory and Its History

The Orthodox View is part of a broader “Orthodox Inventory of the Universe” (Figure 1. This inventory divides the world into two categories: persons (subjects) and things (objects). This division is understood to be of the highest level (summa divisio) and exhaustive (tertium non datur; “there is no third”). If one accepts such an Orthodox Inventory, one must also accept a formal definition of legal thinghood: everything that is not a legal person, is a legal thing.Footnote ³¹

Figure 1

The Orthodox Inventory of the universe. Persons hold rights and/or duties, or the capacity to hold rights/duties. Everything else falls under things

The history of the Orthodox View is often understood to date back to Roman law and the works of the jurisconsult Gaius. However, this is most likely a misconception. Gaius did indeed divide his Institutes (Institutiones) into three parts, addressing persons, things and actions.Footnote ³² But it is unlikely that he would have meant this pedagogical division as presenting three, mutually exclusive categories. The “law of persons”, for instance, does not address whether one is a person or not, but rather the law of personal statuses: whether one is Roman or non-Roman, head of the family or not, and so on. Another telling example is slaves: though typically understood as paradigmatic examples of human beings who are not legal persons but rather things, Gaius mentioned slaves under both persons and things.Footnote ³³

Rather than truly dating back to Roman law and Gaius, the Orthodox Inventory can be understood as a product of modern-era legal scholarship, starting with Renaissance Humanists such as Hugues Doneau (Hugo Donellus, 1527–91) and developed further by Continental European legal scholars. The idea of personhood and thinghood as strictly mutually exclusive categories was prominent in the legal and moral philosophies of Immanuel Kant and GWF Hegel, and the most prominent legal-doctrinal expression of this idea was likely by the great German legal scholar Friedrich Carl von Savigny. These Continental ideas were imported to the Anglophone world by the great English legal philosopher John Austin, who had studied at the University of Bonn, Germany.Footnote ³⁴

Criticism of the Orthodox View

Though the Orthodox View is the prevalent way of understanding legal personhood, it can be criticised. Let us now consider some problems of the view.

The first problem has to do with simplistic nature. Though simplicity is of course a desideratum of a theory, it should not come at the cost of explanatory power. For instance, the Orthodox View struggles at explaining grey areas, such as the just-mentioned legal status of slaves, who were legal nonpersons for most purposes, but persons for the purposes of being liable under criminal law; and gradual changes, such as the steady progress of women towards full legal personhood.

A somewhat more complex issue has to do with what may be termed the discrepancy problem.Footnote ³⁵ The Orthodox View leads to contradictions if we combine it with (1) typical modern theories of rights and (2) certain widely accepted extensional beliefs, that is, beliefs regarding who or what is a legal person. For instance, nonhuman animals today, and slaves in the antebellum US, are both typically classified as legal nonpersons. However, it is highly plausible to think that animals hold, and slaves held, rights. Nonhuman animals are protected in ways that are quite straightforwardly classifiable as rights under the interest theory of rights – though these protections only meet the criteria for “simple rights” rather than “fundamental rights”, under the taxonomy of Saskia Stucki.Footnote ³⁶ Slaves were protected in similar ways, but they also held some legal positions classifiable as rights under the will theory, such as the right to appeal criminal convictions.Footnote ³⁷ Hence, the Orthodox View, modern theories of rights and said extensional beliefs are incompatible.

The Capacity-for-Rights view, in particular, suffers from problems having to do with its origins in private-law theory. The notion of giving some X the “capacity for rights and duties” has explanatory power within private law: private law has to do with setting up a framework where individuals and other actors can acquire rights and duties as a result of transactions and delicts. However, this idea of a capacity for rights does not make as much sense outside of private law. For instance, if we take certain criminalisations – such as the crimes of battery and murder – to entail rights (as the interest theory would have itFootnote ³⁸) then the legislator needs to simply enact such criminalisations in order to give these kinds of rights to X. There is no need for a separate declaration that X “can hold rights within criminal law”; such declarations are superfluous. Some other declarations – such as endowing X with legal standing and the status of a criminal-law victim – are of course meaningful, but merely conferring upon them the “capacity for rights” is not. Conceptualising legal personhood merely in terms of the capacity for rights would also lead to some strange consequences. Let us suppose that Solveig could enter contracts and own property, but that anyone could also freely hit or kill her without any consequences. It would at least be somewhat strange to label her a legal person. It is more plausible to understand legal personhood as partly about the “capacity” for certain rights and/or duties, but partly as the actual holding of certain rights and/or duties.

Monism and Pluralism

MonistFootnote ⁵⁸ accounts emphasise the connections between personhood simpliciter and legal personhood, whereas pluralist accounts emphasise the distinctiveness of legal personhood from personhood. I take most realist accounts to be committed to monism in some form or another. The connection posited by monism can, prima facie, be understood in at least three different ways.

First, monism can be understood as a thesis about who or what can be a legal person. The most straightforward thesis here would be that persons, and only persons, can be legal persons.

Second, we can take monism to be a thesis about who or what should be a legal person. This is how Benjamin Sachs interprets Naffine’s realism.Footnote ⁵⁹ Again, the most straightforward way of understanding this thesis would be that persons, and only persons, should be legal persons.

Third, monism can be understood as a conceptual thesis. For instance, Michael Moore seems to think that there is no distinct legal personhood at all, but rather simply personhood. He takes personhood in legal contexts to be a “borrowed” concept, though he does not specify what he means by this.Footnote ⁶⁰ According to pluralism, on the other hand, personhood and legal personhood are distinct concepts.

The “can” and “should” questions will be addressed in the sections below, but I will briefly discuss the conceptual thesis here. What exactly does it mean that law would be working with a “borrowed” concept? Law can of course employ non-legal concepts. However, when doing so, the law needs to attach legal consequences that follow if the criteria of the concept obtain. For instance, let us assume that the prohibition of “cruel and unusual punishment” in the US Constitution involves the use of a borrowed moral concept, namely that of cruel. However, the legal consequences of when a punishment is deemed cruel and unusual are set out in the law: such punishments are prohibited by the Constitution. The same can be said of (legal) personhood: even if personhood were a borrowed concept within law, the legal consequences of something’s being a person would regardless need to be determined. These legal consequences would, then, amount to the status of legal personhood. Regardless, even the status of legal personhood can be derivative from, say, morality. The most straightforward account of such a connection would likely be the following: the status of legal personhood amounts to the legal status that moral persons, according to morality, should have.

Legal Personhood as Status and Construct

Naffine’s category of legalism hides yet another conflation. This is apparent in her treatment of Christopher Stone’s work. She takes him to be a legalist, and goes on to write:

Here, we encounter yet another issue underlying the conceptual-scheme understanding of legal personhood: legal personhood can be understood as a status or a construct.

Understood as a status, legal personhood is a legal property that attaches to certain entities, of which the paradigmatic example is human beings.Footnote ⁶² Law can give various statuses to human beings, such as that of a husband or a convict. When a slave was freed, his or her legal status was changed from that of a slave (and, subsequently, property) into that of a person.

If legal personhood is understood as a construct, legal persons are seen as functioning purely in the legal realm: rather than giving pre-existing entities a legal status, law creates entities.Footnote ⁶³ Such an approach to legal personhood finds most appeal when we consider the way corporations and other similar legal arrangements can be created and used in imaginative ways: corporations can create dozens of subsidiaries for tax purposes and transfer assets and liabilities between them at will. Here, new “legal persons” are created, manipulated, and dissolved for various technical reasons. In these kinds of situations, it is tempting to think that nothing “becomes” a legal person when Company Inc. sets up a subsidiary for holding purposes. A legal person appears here as a construct – a legally created entity – rather than as a pre-existing entity newly endowed with a legal status.

I have sought to make sense of this systematically ambiguousFootnote ⁶⁴ nature of legal personhood by distinguishing between a legal person and a legal platform. A legal person is an entity that is endowed with the status of legal personhood. Such entities can be individual human beings and human collectivities; and potentially for instance nonhuman animals and robots. A legal platform, on the other hand, is merely a legal construct – a bundle of rights and duties with specific features.Footnote ⁶⁵ For every natural person, there is at least one legal platform that attaches to them: for Jane Smith, the platform’s name would be “Jane Smith”; it is under this legal identity that she can operate from cradle to grave. However, in jurisdictions that allow for one-person corporations, she could also set up a separate legal platform, such as “Jane Smith, Inc.”. She could then operate under two legal identities. What is distinctive of the latter platform – as is typical of corporate platforms – is that said platform can relatively freely change its object of attachment: Jane can for instance sell her corporation to someone else. Hence, legal personhood is a status, whereas a legal platform is a construct. Untangling this systematic ambiguity will have some implications for the question of who or what can be a legal person, which will be addressed below.

The “Should” Question in Law: Natural Law and Legal Positivism

From a legal point of view, how the “should” question is addressed depends greatly on what school of legal thought one supports: natural law or legal positivism.

Natural law is the minority approach. Its supporters typically emphasise that one cannot detach law as it is from how it morally should be.Footnote ⁸¹ John Finnis has presented a natural-law thesis about who or what should be a legal person. I will now quote at length his discussion of three US Supreme Court cases where the Court had essentially denied personhood under the US Constitution to black people and foetuses, whereas it had held that corporations are persons under the Constitution:

Finnis here propounds a thesis that he takes to apply to all legal systems. He also seems to subscribe to a monist view, according to which persons – and perhaps only persons – should be legal persons.Footnote ⁸³

A somewhat similar account can also be discerned in the so-called Philosophers’ Brief, written by a number of philosophers originally in support of a US lawsuit demanding legal personhood for two chimpanzees. In the brief, the authors argue for why (certain) animals should be viewed as persons, implying that such an argument would then have legal relevance. Hence, the brief seems to proceed from the assumption that if certain animals are indeed persons, then this fact constitutes a reason for endowing them with legal personhood.Footnote ⁸⁴

On the other hand, a legal positivist will likely treat legal arguments as system-contingent: the validity of claims such as “all persons should be legal persons” will depend on the legal system in question. Similar arguments as those presented by Finnis and in the Philosophers’ Brief might still be relevant in some legal systems. A positivist could argue, for instance, that a legal system is committed to some moral principles which, correctly understood, would extend legal personhood to entities that are not currently recognised as legal persons. For instance, consider a legal system that has in its constitution committed to justice and equality but that has not recognised animals as legal persons. One might then present the argument that the principles of justice and equality, correctly understood, give the judiciary legal reasons to endow (some) animals with legal personhood. Relatedly, if the statute of some country simply makes reference to “persons”, then philosophical arguments of the kind “X is a metaphysical person, therefore X should fall under the ambit of the provision” may very well be relevant. However, in general, a positivist approach tends to understand the question of who or what should be a legal person primarily as a question falling under morality, politics or other such fields.

Let us now look closer at how legal personhood may function in legal reasoning.

Children as Legal PersonsFootnote ¹⁰⁰

Adults who are deemed to be of sound mind are, in Western legal systems, fully active legal persons. Thus, they may normally dispose of their rights and duties, enter into contracts freely and so on. Furthermore, they are held fully responsible under tort law and criminal law, as well as under other types of responsibility. I call this type of fully active legal personhood independent legal personhood.

On the other hand, infants are, under prevalent Western law, what I term purely passive legal persons: they do not bear any legal responsibilities nor can they exercise legal competences. This full passivity may run counter to how some scholars think of the matter. For instance, the German legal scholar Friederike Wapler starts a handbook chapter by simply asserting that “[c]hildren can have rights and duties, such as being party to a sales contract, or purchasing or inheriting property” (emphasis added).Footnote ¹⁰¹ Wapler then goes on to discuss the notion of children’s rights critically but does not address the question of obligations further.

I take the idea that infants bear duties to be motivated by the following reasoning. First, ownership and contracts do typically entail duties. Owning a piece of property may, for instance, incur an annual property tax. The question is, then, whether infants can own property and be parties to contracts. Western jurisdictions employ two primary strategies here. First, a solution favoured by many common-law countries is to use the notion of a trust: a piece of property owned by a child is, formally, understood as being part of a trust, with the child as the beneficiary. Civil-law countries, on the other hand, generally lack the notion of a trust. In many civil-law countries, children are instead understood to own the property directly. However, they only have the right to “enjoy” the property, rather than the right to “exercise” or “administer” it. This arrangement relates to the abovementioned civil-law concept of legal capacity, which is often defined in terms of the capacity to hold private-law rights and duties. Wapler is clearly referring to this doctrinal construction.

Infants can own property and be parties to contracts, and these legal institutions do typically entail duties. Can infants therefore bear duties? Many civilian property lawyers would likely say that the duty is the infant’s, even if the infant is not in practice expected to fulfil the obligation. The main problem with this explanation is obvious: the infant is not really the party who is expected to do anything. Rather, this task falls upon the infant’s representative. Hence, the explanation relies on a problematic notion of a duty.

I have argued that this kind of a situation is best understood as a triangular situation, depicted in Figure 3. Let us assume that Lauri is represented by Maria, who acts as his guardian. We should now distinguish three elements: Lauri, Maria, and Lauri’s legal platform, i.e. the legal rights and duties of Lauri. Given that Lauri is an infant and therefore a purely passive legal person, he cannot administer his legal platform: he cannot exercise his rights or fulfil his duties. Instead, Maria administers the legal platform, and such administration is done in Lauri’s name. Lauri is, however, the beneficiary of the platform: if there are, say, property rights in the platform, it is Lauri who should benefit from them, and Maria has the fiduciary duty to administer the legal platform. There are duties in Lauri’s legal platform, which are currently borne by Maria. The bearer of these duties can change if Maria ceases to be Lauri’s legal guardian. To reappropriate a metaphor of John Austin, Lauri can thus be understood as the ‘compass’ of the duty, who stays constant even if his representatives might change.Footnote ¹⁰² Lauri may also have several representatives who represent him in different capacities. For instance, Finnish law distinguishes between a child’s custodians and guardians.Footnote ¹⁰³ A more comprised way of expressing the same idea could be to label these duties as “proxy duties”, signifying the idea that any talk of the “duties” of Lauri is really a shorthand for the duties of Lauri’s proxy, whoever that proxy might be.

Figure 3

The triangular relationship between a passive legal person, their legal platform, and their representative

To sum up, neonates do not bear legal duties and should therefore not be understood as active legal persons. However, children leave full passivity at a relatively young age, as they gradually begin to be treated as legal actors. The treatment of children as active participants and agents – rather than merely as objects of protection – has increased significantly since the adoption of the Convention on the Rights of the Child in 1989.Footnote ¹⁰⁴ Even relatively young children typically have some control over their rights and duties and can be held responsible to some extent. They are thus not fully passive, yet not independent either. This position can be labelled ‘dependent active legal personhood’, or dependent legal personhood for short.

This dependent legal personhood can take somewhat different forms. The “first stage” of dependent legal personhood is the right to be heard, which is held already by young children according to the Convention on the Rights of Child. However, the right to be heard does not mean that one’s views determine the outcome; rather, under the convention, the views of the child shall be given “due weight”. A legal competence, on the other hand, is the capacity to effect an intentional legal change. Legal competences are thus determinative and conclusive. The child gains access to an incident of active legal personhood par excellence when she is endowed with her first competences.

As regards the legal competences of children, we can distinguish three main categories (see Table 4). First, a negative competence is the power to prevent a legal change, initiated by another. Second, dependent competences are subject to a consent, veto or cancellation by some other party. Finally, a competence is independent if its holder may freely decide to exercise it, and the legal effect is not subject to, for example, a consent by some other party.Footnote ¹⁰⁵

I will use the Finnish Act on the Freedom of Religion as an example because it contains an interesting provision that exemplifies these different categories.Footnote ¹⁰⁶ As regards membership in a religious community, Section 3(1–3) sets out the following rules:

The provision thus contains four different categories:

1. (1) Children under 12 years old, whose religious affiliation is decided by those who have custody of them.

2. (2) Children between 12 and 17 years, who must consent to a change in their religious affiliation (negative competence, approval).

3. (3) Children between 15 and 17 years, who may initiate a change in their religious affiliation, though requiring the consent of their guardians (dependent competence).

4. (4) Adults, who may freely decide over their religious affiliation (independent competence).

The path of a child from a purely passive legal person to an independent legal person is thus a gradual development. Infants are not entitled to being heard nor are they endowed with legal competences. As the child grows older, she must be heard in an increasing number of contexts, and she will gradually also receive an increasing number of competences. Figure 4 depicts how law’s relationship towards the child gradually shifts from paternalism to recognising the child’s autonomy and independence.

Figure 4

The competences that have been conferred to a child can be seen as an expression of the extent to which the legal system recognises the child’s autonomy and independence, as opposed to taking a paternalistic stance towards the child

Persons with Cognitive Disabilities

Persons with cognitive disabilities have traditionally been denied the competence to decide about their own affairs.Footnote ¹⁰⁷ They have often been unable consent to medical procedures or even to sex.Footnote ¹⁰⁸ However, in the recent years, disability law has seen a somewhat similar paradigm shift as the legal status of children. The agency, autonomy, and self-determination of persons with disabilities is increasingly emphasised, with the Convention on the Rights of Persons with Disabilities spearheading a new approach. Article 12 of the Convention includes the following sections:

1. (1) States Parties reaffirm that persons with disabilities have the right to recognition everywhere as persons before the law.

2. (2) States Parties shall recognise that persons with disabilities enjoy legal capacity on an equal basis with others in all aspects of life.

3. (3) States Parties shall take appropriate measures to provide access by persons with disabilities to the support they may require in exercising their legal capacity.

As noted in Section 1, the terminology here is ambiguous. The phrase “legal capacity” is often understood as the capacity to hold rights and duties. However, “legal capacity” in the Convention denotes instead the capacity to administer one’s own legal situation through legal competences.Footnote ¹⁰⁹ To make matters even more complicated, the phrase is also used to refer to both passive and active legal personhood.Footnote ¹¹⁰ In what follows, I take “legal capacity” to refer to active legal personhood, and in particular to legal competences.

The new approach spearheaded by the Convention, labelled “universal legal capacity”, entails in particular that

This is not to say that persons with disabilities need to make such decisions alone; rather, the individual should receive assistance, such as a circle of support consisting of people who are familiar with the individual. However, the individual may at any time reject the support. This arrangement is denoted supported decision-making.

On the other hand, “regimes of substituted decision-making” must be abolished. Under such regimes, “legal capacity is removed from a person (even if limited to a single decision) and a substitute decision maker appointed by a third party takes decisions based on what he or she considers is in the best interests of the person concerned, even if that goes against the will of the latter”.Footnote ¹¹² The Convention thus gives precedence to the will, rather than the best interests, of individuals with disabilities.

In conclusion, the new regime introduced by the Convention requires treating persons with disabilities increasingly as active legal persons.

A Very Brief History of Corporations

To understand corporations, we should briefly consider their history in Western law. I will now briefly discuss the history of corporations in general as well as the origins of modern business corporations and of the US doctrine treating corporations as constitutional persons.

The notion of a corporation – in the broad sense of the term – dates back to Roman law. The first Roman corporations (universitates, singular universitas) were public bodies such as cities and states. They could own property and enter into contracts in their own name. It has been suggested that the notion of a corporation spread throughout Europe with the Roman administration.Footnote ¹¹⁷ In the Middle Ages, the notion of monasteries’ owning property was important, as for instance monks were not allowed to own property. As a result, the canon-law status of monasteries and other corporations needed to be specified. Famously, Pope Innocent IV declared them to be fictional persons (personae ficta). Some norms applying to human beings, such as the possibility of excommunication, did not apply to fictional persons, but they could regardless have their own assets.Footnote ¹¹⁸

The earliest examples of modern business corporations were the English East India Company and the Dutch East India Company, founded in 1600 and 1602, respectively. These companies exhibited at least one feature of modern business corporations: they were joint-stock companies. However, it is disputed whether these companies already exhibited the second crucial feature of modern business corporations: the limited liability of shareholders. Some scholars argue that they did, whereas others place the emergence of limited liability at the 1850s or even as late as the 1930s.Footnote ¹¹⁹

The emergence of the US doctrine of the constitutional personhood of business corporations is usually traced back to how the US Supreme Court interpreted the Equal Protection Clause of the 14th Amendment, which states that no state shall “deny to any person within its jurisdiction the equal protection of the laws”. In the 1880s, the Court declared that corporations are “persons” within the ambit of said clause.Footnote ¹²⁰ The implications of the constitutional personhood of corporations have been subject to a considerable body of case law; one of the most famous recent cases is Citizens United, which declared restrictions on election campaign spending by incorporated entities as unconstitutional on the grounds that they violated the free speech of corporations.Footnote ¹²¹

Theories of Corporations

A scholarly discussion of corporations typically proceeds from a relatively canonical list of so-called corporate theories. Often, at least three kinds of theories are listed: the association (or aggregate or partnership) theory, the grant (or concession or fiction) theory and the real-entity theory.Footnote ¹²² The association theories explain corporations in a “bottom-up” manner: companies consist of individuals coming together, and the interests of corporations are, actually, the interests of said individuals. Grant theories explain corporations as creations of the law. Finally, real-entity theories typically emphasise the “pre-existing” nature of corporations: they are collectives that “really” exist, and the law simply provides recognition of this fact.Footnote ¹²³

However, one should be careful here. First, it is not always obvious whether these theories are mainly offered as theories of business corporations or rather of incorporated entities more broadly. Second, merely listing different corporate theories – often presented, at least implicitly, as mutually exclusive alternatives – can even be misleading. For instance, David Millon analyses corporate theories in terms of three dimensions:

1. (1) entity/aggregate: “the corporation as an entity, with a real existence separate from its shareholders and other participants, and the corporation as a mere aggregation of natural individuals without a separate existence”;

2. (2) artificial/natural: “the corporation as an artificial creation of state law and the corporation as a natural product of private enterprise”; and

3. (3) public/private: corporations have “broad social and political ramifications that justify a body of corporate law that is deliberately responsive to public interest concerns”, as opposed to “corporate law as governing little more than the private relations between the shareholders of the corporation and management, which acts as their agents”.Footnote ¹²⁴

We can see here that, for instance, the real-entity and aggregate theories will differ on the first axis, but they may very well agree on the second axis: both will likely see the (business) corporation as arising out private enterprise.

Different theories may also seek to explain different aspects of corporations. Petri Mäntysaari distinguishes between legal and economic theories of corporations.Footnote ¹²⁵ The abovementioned theories are typically legal theories, and more specifically doctrinal theories: they often seek to explain what rights and duties corporations have, or should have, as a matter of the law of some jurisdiction. Hence, for instance, grant theories emphasise that corporations have only those rights and duties that have been explicitly granted to them. Some theorists have also emphasised that limited liability is a type of “privilege” granted to limited liability companies – a privilege that also justifies additional duties imposed upon such companies.Footnote ¹²⁶ Associate theories, on the other hand, may quite naturally favour a broader set of rights and/or duties, because the rights of corporations are an extension of the rights of the participants.Footnote ¹²⁷ On the other hand, philosophical theories of corporations – recently often propounded as part of the emerging field of social ontology – typically seek to provide an analytical and descriptive explanation of corporations. I will now briefly discuss two issues from this analytical point of view: the social-ontological nature of corporations, and the nature of the legal act of incorporation.

Analytical Questions

When addressing what corporations are – rather than, for example, what legal rights they should have – I take the most appropriate labels for the alternatives to be realism and fictionalism.

A realist approach takes collectivities to “really” exist, independent of their legal status: they are, for instance, conceived of as group agents, meaning organised groups that – as a result of the organisation – exhibit agency that is clearly distinct from the agency of the individual members. What incorporation means, then, is giving these collectivities a specific legal status. I have elsewhere used the metaphor of a “visibility cloak”: incorporation renders group agents visible within the law, enabling them, for instance, to perform legally enforceable transactions.Footnote ¹²⁸

A fictionalist approach, on the other hand, provides a different descriptive account of the corporation. According to this approach, corporations are examples of the kinds of imaginary constructs that law can create: corporations exist “only in contemplation of law”, as the US Supreme Court has put it.Footnote ¹²⁹ Fictionalism can in fact take two rather different forms, having to do with the ambiguity of the term “fiction”. In the narrower “as-if” sense, “fiction” refers to the law pretending that something is the case.Footnote ¹³⁰ A classic example of a fiction theory in the “as-if” sense is the theory of Friedrich Carl von Savigny. According to Savigny, only human beings are “real” persons because they have the kind of inner freedom that is requisite of persons.Footnote ¹³¹ The idea of a corporation, therefore, has to do with creating an artificial person and thereby treating non-persons as persons for legal purposes.Footnote ¹³² We can see that Savigny is here positing a connection between personhood and legal personhood, and thereby assuming a monist position about legal personhood (see Section 2).

In a broad sense, “fiction” means simply “social fact”. A social fact is, roughly put, a fact that depends on the beliefs and attitudes of some social group. For instance, the existence of money is a social fact: whether some piece of fabric is money depends on whether we collectively treat it as such.Footnote ¹³³ In this broad sense, corporations are almost indisputably fictions: corporations’ existence depends on various beliefs and attitudes of human beings.

However, some scholars emphasise how corporations can be created “out of thin air”. For instance, John Searle argues that when a new corporation is formed, nothing becomes a corporation:

For Searle, a corporation is “free-standing”, meaning that nothing counts as a corporation. Hence, corporations are not pre-existing entities – such as group agents – that are given a legal status, but rather simply legal creations. In the terminology introduced in Section 2, corporations are constructs rather than statuses.

I believe that the Searlean and the realist pictures can be combined. As noted above, the word “corporation” is occasionally used to refer to an entity with a legal status, and occasionally to the status (or construct) itself – to a bundle of rights and duties. In this latter sense, there is nothing that is made into a legal person; a new “legal person” simply comes into being. This sense of legal-person-qua-legal-positions becomes most apparent when we consider so-called single-person corporations, possible in some jurisdictions.

Let us say that Mary starts a veterinary practice. She wants to keep her business-related assets and liabilities separate from her private assets and liabilities, so she starts a one-person corporation, Mary Inc., for her business. She now has two personas, “legal masks”, through which to sign contracts, own property and so on: “Mary” and “Mary Inc.”. These two masks are just two bundles of legal positions (rights and duties), both attaching to Mary. Corporations qua bundles of rights and duties may be called corporate platforms,Footnote ¹³⁵ whereas a group agent with the legal status of a corporation may be denoted an incorporated collectivity.

Exposing this ambiguity about corporations can also serve towards clarifying certain puzzles in social ontology. For instance, let us reconsider Searle’s position. According to Searle, a corporation is “free-standing”, meaning that nothing counts as a corporation. Clues such as Searle’s characterisation of the corporation as “power relationships between actual people” suggest that he seems to be thinking of corporations as corporate platforms.Footnote ¹³⁶ Many critics of Searle have disputed this understanding of corporations. Frank Hindriks argues instead that organisations such as corporations are “constituted by one or more persons”Footnote ¹³⁷, and that “the members count as an organisation with a particular status”.Footnote ¹³⁸ Hindriks, thus, understands corporations as incorporated collectivities. The two can therefore be said to be using the term in different senses.

“Corporation” can thus refer both to corporate platforms and to incorporated collectivities. Some disagreements in corporate theory can potentially be understood in terms of differing explananda: theories that treat corporations as mere legal creations seek to explain corporations qua legal platforms, whereas theories that accord corporations a more robust status understand them as group agents with a specific legal status.

To wrap up our discussion of artificial persons, we can note that even seemingly incompatible theories may coexist to some extent. For instance, consider theories seeking to explain the ontology of corporations at a general level vis-à-vis theories addressing the constitutional status of corporations. One might hold that corporations really exist as collectives, and regardless think that merely this fact does not entitle them to any constitutional rights. Instead, one might employ the association theory to assess whether corporations should be treated as holders of some constitutional rights: essentially, any constitutional rights of corporations should be justified by the constitutional rights of their shareholders.

Animal Status de lege lata

There is, in fact, considerable disagreement on the current state of affairs: whether animals already are legal persons and/or right-holders, and what exactly would be needed to turn them into such.

We can again distinguish a formal and a substantive approach to the legal status of animals. The binary, formal view – often based on the Orthodox View of legal personhood (and, more broadly, the Orthodox Inventory) – understands animals currently as (1) property, (2) nonpersons, and (3) without rights. Furthermore, on this view, turning animals into legal persons would typically require some kind of an explicit, formal act, such as a declaration by a court or a legislature stating that animals are legal persons or that they have rights. On the other hand, a substantive view offers a different way of approaching the matter. On a substantive view, animals could already hold legal rights while remaining as property, and their legal status could gradually shift towards legal personhood.

Discussions of animal legal personhood have mostly been dominated by the formal Orthodox View. Most theoretically-oriented animal law scholars have assumed that, in typical Western jurisdictions, animals are neither legal persons nor holders of legal rights. One influential legal-theoretical view has been that of Steven Wise, who has in his works assumed that animals must first be declared legal persons, with the capacity to hold rights, before they can actually hold rights.Footnote ¹⁴⁵ His view thus falls under what I call the Capacity-for-Rights view, a version of the Orthodox View. The Capacity-for-Rights view still structures jurists’ thinking about animal rights; it is likely especially powerful among civil lawyers, who are in their legal studies often taught the Capacity-for- Rights view as the definition of legal personhood.

However, not all scholars and judges follow the approach just outlined. First, some have argued that animals, in fact, already hold rights. For instance, Cass Sunstein and the federal judge in the Tilikum case have both stated that animals already have rights as a result of animal welfare laws.Footnote ¹⁴⁶ I myself have put forward a similar argument, as have for example Clare McCausland and Alasdair Cochrane.Footnote ¹⁴⁷ Saskia Stucki has presented the useful distinction between simple rights and fundamental rights. According to Stucki, animals could be said to hold simple rights as a result of animal welfare laws, but not fundamental rights, meaning strong, broad and properly enforceable rights.Footnote ¹⁴⁸

Another interesting example of animals’ potentially already holding rights is the issue of pet trusts. Some US states already allow for pet owners to leave some money in a trust, with their companion animal as the beneficiary. Given that such arrangements are typically understood as ownership, one may in fact argue that companion animals can already own property. Hence, such pets have, in fact, already been endowed with at least one incident of legal personhood – and a significant incident at that.Footnote ¹⁴⁹ However, though such arrangements constitute a rather significant departure from how animals have traditionally been placed in the person/thing dichotomy, these arrangements can also be seen as problematic from an ethical point of view, given that they prioritise certain animals that humans have bonded with, rather than giving rights to animals on an equal basis.Footnote ¹⁵⁰

Scholars departing from the formal paradigm have also outlined alternative ways towards animal legal personhood than the path of formal declaration. Eva Bernet Kempers has argued that the potential progress of animals towards legal personhood could also happen “bottom-up”, as it were, through the gradual improvement of animals’ legal status, whereby legislative declarations of animal personhood would mean formally enshrining a position that has already been developing under the surface.Footnote ¹⁵¹

Proposals for Animal Legal Status

Another question is what kind of a legal status animals should have. Here, scholars have made a number of proposals, which can roughly be categorised into two: those remaining within the person/thing dichotomy, and those departing from it.

First, some have advocated for the legal personhood of animals – though what they mean by “legal personhood” varies. For Wise, legal personhood means the capacity for rights, so merely declaring animals to be legal persons would not amount to much; however, the Nonhuman Rights Project has championed the fundamental right to liberty enshrined in the writ of habeas corpus for some animals.Footnote ¹⁵² Under the abolitionist approach, championed by Gary Francione, animals’ status as property would be abolished completely, and they would receive strong fundamental rights.Footnote ¹⁵³ Some other scholars writing in the common-law tradition have proposed arrangements that combine legal personhood with property status. Examples include the living property arrangement envisaged by David Favre and Angela Fernandez’s proposal for a “quasi-property” status for animals – though Fernandez’s proposal can also be seen as an analysis of the current legal status of animals.Footnote ¹⁵⁴

Civilian jurists have also often approached the issue in terms of a new doctrinal personhood category. For instance, Stucki suggests the category of “animal persons” (tierliche Personen).Footnote ¹⁵⁵ Caroline Regad and Cédric Riot have made a relatively intricate proposal. They suggest that the category of natural persons be divided into human persons and non-human persons. Furthermore, they divide the latter category into three subcategories: companion animals, other domesticated animals and wild animals.Footnote ¹⁵⁶ Each subcategory would come with a distinct set of rights.Footnote ¹⁵⁷

A second approach involves instead at least a partial departure from the person/thing dichotomy. Tomasz Pietrzykowski, for instance, has argued that animals be granted the status of a non-personal subject of law, with the right to have their interests considered.Footnote ¹⁵⁸ Maneesha Deckha, on the other hand, has argued for the third category of “legal beingness”.Footnote ¹⁵⁹

The Meaning of AI Legal Personhood

The legal personhood of animals and nature is typically motivated by the protection of these entities. However, AI personhood may be motivated by other reasons as well. Simon Chesterman writes:

Chesterman also notes – as discussed earlier in this Element – that granting legal personhood to an AI system can be based on intrinsic or instrumental reasons. Reasons for granting legal personhood would likely be reflected in the type of ensuing legal personhood. I myself have distinguished three contexts for AI legal personhood: the ultimate-value context, commercial context and the responsibility context. These three contexts can, however, overlap to a significant extent, as will become apparent.

The ultimate-value context is quite close to what Chesterman means by “intrinsic reasons”. Essentially, AI systems would be granted incidents of legal personhood for their own sake. The appropriate type of legal personhood arrangement would depend on the type of entity we are talking about. If the AI system were a robot with a bodily form, it could be granted rights protecting its physical integrity. Its autonomy and self-determination could be protected by various incidents of legal personhood, such as ownership, contracting and legal standing. The ultimate-value context is not clearly reflected in Chesterman’s typology of “who to blame when things go wrong” and “who to reward when things go right”. Rather, the question here is, “who or what is (morally) entitled to legal personhood”. Theoretically, even purely passive legal personhood could be possible here.

What I have termed the commercial context has to do with AI systems as actors in the commercial sphere. Here, we do not merely consider robots as agents that can cause damage, but as agents that can for instance enter contracts, buy and sell stock and so on – in short, perform transactions through the use of legal competences. Another interesting question is whether an AI system could own the copyright for an artwork it has produced, or whether it could successfully file for a patent for an innovation.Footnote ¹⁶⁹ This question can be seen as pertaining to AI ownership rights in a broad sense.Footnote ¹⁷⁰ Again, questions such as these can be approached from an intrinsic or instrumental point of view. From an intrinsic point of view, we may think of the commercial context as being derivative of the ultimate-value context. Hence, for instance, AI systems’ intellectual property rights could be justified by reference to their (moral) right to be recognised as makers of artworks created by them.Footnote ¹⁷¹ However, again, the instrumental point of view is more prominent.

A relatively established way of approaching these questions – especially with regard to contracting and other transactions – has been to employ the tool/representative/legal person trifurcation.Footnote ¹⁷² As the terms imply, treating an AI system as a tool would imply treating it like a hammer; the second category would entail applying the law of agency to, for example, transactions entered by an AI on behalf of, say, a company; and the third would involve treating the AI as a separate legal person. However, I have argued that this trifurcation in fact conflates two important aspects: (1) the extent to which the AI’s legal platform – bundle of rights and duties – is treated as separate from some other legal platform (separateness) and (2) the extent to which the AI system is treated as a distinct actor in the eyes of the law (independency).Footnote ¹⁷³

Though I have earlier offered this analysis in the context of AIs as commercial actors, I now believe the question of the active legal personhood of AI systems more generally – including the responsibility context – can be approached along these lines. However, addressing this point will require first presenting the responsibility context.

The responsibility context is mainly focused on questions of responsibility when things go wrong. Given the increasing autonomy and sophistication of robots and other AI systems, traditional doctrines of how to attribute blame for harm might not be appropriate. The responsibility context may also be understood from an intrinsic or instrumental point of view. From an intrinsic point of view, we might ask whether a robot could be the kind of entity that deserves to be held responsible, or even benefits from it. For instance, treating a robot as responsible could be seen as supporting the development of the robot as a moral agent. However, the responsibility of AI systems is more often approached from an instrumentalist point of view, as performing some societal function. A central issue here has to do with the so-called responsibility gap: “the risk that no human agent might be legitimately blamed or held culpable for the unwanted outcomes of actions mediated by AI systems”.Footnote ¹⁷⁴ One solution to this issue may be holding the AI system itself responsible. We can distinguish compensation and punishment here.

First consider the idea that AI systems be required to pay compensation for some of their actions. An oft-mentioned example is that of autonomous cars: if an autonomous car causes damage, could it be required to compensate for the damage itself?Footnote ¹⁷⁵ Such an arrangement would be one solution to the responsibility gap, though many other solutions have been proposed as well, such as insurancesFootnote ¹⁷⁶ and “no-fault liability schemes”, meaning statutory compensation schemes that are used to compensate for losses even when no-one is at fault.Footnote ¹⁷⁷ Making the AI system itself liable to pay compensation would necessitate that the AI system also be able to own property and be party to transactions, and would therefore take us closer to the commercial context.

Another way of possibly filling the responsibility gap would be to punish an AI system. Some have argued that such a punishment could even serve the human psychological need for revenge.Footnote ¹⁷⁸ An important antecedent question here is whether artificial intelligence systems even can be punished. For instance, Robert Sparrow has famously argued that the use of autonomous weapons systems may lead to a situation where no-one and nothing can be held responsible for, say, civilian deaths. According to Sparrow, the use such of systems leads to a genuine responsibility gap because it is impossible to punish the weapons systems themselves.Footnote ¹⁷⁹ Many scholars, including myself, have argued that it is indeed meaningful to punish AI agents: even if they might not be meaningful objects of punishment from a retributivist point of view, other rationales of punishment, such as deterrence, may very well apply to AI agents.Footnote ¹⁸⁰ Possible sanctions would involve economic sanctions – especially if the AI system could own property – as well as other sanctions, such as “forced labour” (using the AI system for a specific purpose), reprogramming or even destruction of the system in question.Footnote ¹⁸¹

Now, we can also distinguish the independency of responsibility from the independency of competences. This dimension tracks the question of whether harms caused by an AI system are attributed to some other party (such as its owner) or to the AI system itself. In a way, the question here is “where the buck stops”. If the AI is treated as fully responsible, the buck almost always stops at the AI: harms for which it is causally responsible are, as a rule, not attributed to any other party.Footnote ¹⁸² However, there can still conceivably be similar kinds of exceptions as with the responsibility of human beings: for instance, an adult of sound mind who is in most regards fully responsible, can be misled into performing acts whose consequences they are not fully aware of. In such cases, the responsibility may ultimately lie with the party doing the misleading.

Under partial responsibility, the buck occasionally stops at the AI, but there may also be cases where some other party is held responsible for the acts of the AI. For instance, the programmers of the AI might be held responsible in some cases. Finally, under assimilation – as with the independency of competences – the AI system is essentially not treated as legally capable of performing acts at all. Like in the commercial context, whether the AI has a separate legal platform affects many issues as well. For instance, as discussed above, requiring the AI pay compensation would necessitate that the AI also controls a legal platform. These different categories are presented in Table 5.

Are Some AI Systems Already Legal Persons?

As has hopefully become clear, the question of AI legal personhood is multifaceted and multifarious. Hence – as is typical of legal personhood – the question of whether some AI systems can, are, or should be, legal persons will depend on the context.

First, can AI systems be legal persons? If one adheres to the Anything-Goes View, the answer is of course straightforward: yes, they can. If the legislator or some other appropriate legal authority decrees that some AI system is a legal person, then it is a legal person.Footnote ¹⁸³ If one instead thinks that not everything can be a legal person, some account of the domain of legal personhood is required. I myself have argued that these limits can be determined by considering the building blocks of legal personhood: claim-rights, duties and competences, and which entities can be endowed with said building blocks.Footnote ¹⁸⁴ However, I will not delve deeper into this question here.

A rather interesting question is whether some AI systems are already legal persons. There are at least two prominent potential examples. First, the robot Sophia, which was supposedly granted citizenship by Saudi Arabia.Footnote ¹⁸⁵ However, the legal implications of this putative citizenship – if there are any – are extremely unclear. As the journalist Robert David Hart puts it,

In lack of evidence to the contrary, I take this case as nothing more than a marketing stunt.

The second, and much more interesting, example is the legal regulation of personal delivery devices, meaning small robots that deliver goods from a store to a consumer. As scholars such as David Gunkel have noted, the legislation of the state of Virginia provides that “a personal delivery device operating on a sidewalk or crosswalk shall have all the rights and responsibilities applicable to a pedestrian under the same circumstance”.Footnote ¹⁸⁷ Gunkel claims that the legislature has in fact granted rights and responsibilities to said robots. However, this interpretation may be questioned.

Gunkel’s view relies on what I have termed “Rights Deference”: taking legislative ascriptions of rights very literally and unquestioningly.Footnote ¹⁸⁸ However, instead of being this deferential, one can follow Hans Kelsen’s approach. Kelsen notes that a law according to which “a good which is not in time returned to the sender has to be treated as if it had been approved and accepted by the receiver” need not be analysed as claiming that the receiver really has approved the goods. Instead, Kelsen takes the legislative text to be an “abbreviating expression”, noting that “[i]t would be superfluous to repeat all the rules which have already been set down for the first case. The legislator can rest content with declaring that in the second case the same rules apply as in the first case”.Footnote ¹⁸⁹ Even if the Virginian legislator has not used the phrase “as if” – as in Kelsen’s example – one can treat the provision as an abbreviating expression. For instance, according to the same statute, even bicycles have rights.Footnote ¹⁹⁰ Hence, there are reasons to think that – in spite of the wording of the law – personal delivery devices and bicycles might not, in fact, be right-holders or duty-bearers. However, addressing this topic more thoroughly would require an extensive foray into theories of rights.

Regardless of whether we may plausibly reach the conclusion that personal delivery devices and bicycles hold rights and responsibilities under Virginian law, this does not yet settle the question of whether they are legal persons. Under the formal Orthodox View, all right-holders and duty-bearers are of course legal persons. However, under the Bundle Theory, the devices in question would not be legal persons, as they lack virtually all of the incidents of legal personhood. For instance, if a driver were to drive over a personal delivery vehicle, they would assumedly not be required to pay compensation to the device itself, but rather to its owner or other such party. Hence, this case is in fact a good example of the problems of the Orthodox View in explaining legal personhood.