1. The Early Cambridge School

The Cambridge School started as a reaction to treating the canon as a timeless conversation conducted across the armchairs of political philosophers. Skinner, for instance, targeted Leo Strauss's advocacy for reading the greats to uncover the true meaning hidden in the texts of the geniuses of political thought.Footnote ⁷ This esoteric but timeless meaning had to be hidden from the masses, Strauss thought, because political genius almost by definition departs from the common political traditions of the day. Philosophers must use rhetorical tricks to disguise their true meaning from would-be persecutors, all the while leaving hints for those with “ears to hear” who can pick up and continue the real acontextual philosophical argument among safe company.Footnote ⁸

Skinner took the rhetorical tricks used in the texts not as a disguise covering their true meaning, but instead as constituting part of that very meaning. As Wittgenstein put it, “the meaning of a word is its use in the language.”Footnote ⁹ Meaning is always on the surface, not hidden. Skinner understood the idea of discovering esoteric meaning as a chance to cherry-pick passages confirming one's political ideology, ultimately leading to a form of confirmation bias.Footnote ¹⁰ When there is a deep meaning that can be grasped by only a few, some can claim they are the few and then cherry-pick the meaning of each text until the canon is nothing more than an echo chamber for their views.

Wittgenstein's account of “meaning as use,” where language is taken as a series of social games with a family resemblance to one another, gave Skinner a contextualist philosophy he thought could overcome such bias. John Locke's influential Two Treatises of Government should not be understood as part of a timeless conversation with Thomas Hobbes's Leviathan; it should be described instead as a series of strategic actions or games designed to advance Locke's own agenda, situated in Locke's particular historical context. The language use in Two Treatises was designed to advance the Whig movement in late seventeenth-century England, not to win a philosophical victory over Hobbes. We know this by virtue of information about Locke's context, such as the influence and patronage of the Earl of Shaftesbury, a noted Whig with lofty ambitions in the Westminster parliament.

Skinner also drew from Paul Grice's theory of meaning to defend this picture, notably that the meaning of an assertion S is the intention an audience believes x by recognizing that this belief was the author's intention in asserting S.Footnote ¹¹ The author conveys this intention with conventions embedded in the common knowledge shared by writer and audience. Given that these conventions are situated in the author's historical and geographical context, unraveling the meaning of a text like Two Treatises of Government must involve identifying the context-bound conventions Locke drew on in England in the late seventeenth century to communicate with his audience. His audience was not Hobbes; Hobbes was dead.

Yet by problematizing the orthodox account, the Cambridge School left itself open to the skeptical challenge: other than for the sheer joy of doing history, why should we study the Western classical canon if it is so wedded to outdated contexts?Footnote ¹² While we may buy Wittgenstein's “meaning as use” slogan, it still raises the question why we should be concerned about the meaning of a classic text in the first place. For better or for worse, appeal to the timeless philosophical conversation was central to justifying the study of the political classics and, without this appeal, it is not immediately obvious why the history of political philosophy matters for contemporary argument.

The Cambridge School's response here is that the canon of political thought still has contemporary relevance because it consists of a series of texts that were historically responsible for the linguistic alterations of our political terminology.Footnote ¹³ Skinner explains, “if we wish to do justice to those moments when a convention is challenged or a commonplace effectively subverted, we cannot simply dispense with the category of the author.”Footnote ¹⁴ Most of the classics of political philosophy were influential in this process of challenging and replacing conventional meanings. Hobbes, for example, was pivotal in radically altering the meaning of terms like “the state” and “liberty.”Footnote ¹⁵ By breaking the boundaries of linguistic convention—“the limits of my language mean the limits of my world”Footnote ¹⁶—the classics ushered in the conceptual frameworks for our political traditions. Philosophical analysis, by this account, amounts to understanding the different uses of language, and the best way to do that is to study the context that motivated the political tradition under scrutiny. Skinner based this interpretation of meaning change on the works of yet another philosopher of language, J. L. Austin. In this view changes in the meaning of political vocabulary are best understood by uncovering the authorial intention of those texts responsible for the change.Footnote ¹⁷

While we agree it is the linguistic innovation of a text that renders its analysis significant, the use of Austin here is problematic and leads the Cambridge School into putting undue emphasis on authorial intention. It is responsible—we conjecture—for the lurking charge of relativism directed at the school and understandably pushes those studying the history of political thought away from contextualism and towards more esoteric interpretations.

Skinner argues we should focus on “what a given agent may be doing in uttering his utterance.”Footnote ¹⁸ This is Austin's concept of illocutionary force,Footnote ¹⁹ where, in addition to the literal semantic meaning of an assertion that can be teased out with Gricean intention, there is a force associated with the action performed in uttering the assertion.Footnote ²⁰ An illocutionary action is something like urging or informing. A good rule of thumb for identifying an illocutionary act is that it makes sense to include the purpose in the assertion itself. It makes sense to say “I warn you to stay at home” or “I urge you to stay at home” (hence warning and urging are illocutionary acts), but not “I manipulate you to stay at home” (so manipulation is not).

Much has been written on whether Machiavelli was intending to mock the Medici family (who had previously tortured and ostracized him) in writing Il Principe. We might say, among other things, that in writing it Machiavelli performed an act of mockery. Mockery was not, however, an action Machiavelli performed by writing Il Principe. The effect of Il Principe’s publication (it is doubtful the Medici ever read it) is of no consequence to whether his work was written as a piece of satire. The effect of publication is, however, important for ascertaining whether Machiavelli performed an act of manipulation, for example, since to know whether Machiavelli manipulated we need to know whether he manipulated somebody. We have to bear in mind the reception of his text and the effect it had on its audience. A study of manipulation is a study not of Austinian illocutionary force but of Austinian perlocutionary force. The difference is subtle: an illocutionary act is what is done in uttering something, whereas a perlocutionary act is done by the performance of an utterance. Because we do not know the effect of our assertion before we have asserted it, unlike illocutionary acts it does not make sense to include a perlocutionary act in the assertion itself. It does not make sense to say “I manipulate you to go home” because whether you have manipulated somebody depends on the effect of the assertion, not the assertion itself.

Now, changing the meaning of political terms in a linguistic community is a decidedly perlocutionary act. It depends on an audience. Whether the meaning of the term “liberty” changes because of a text depends on the effect of the text on an audience. It is something that an author effects by uttering or writing an argument, not in uttering or writing it. The reception of a text describes the perlocutionary action of linguistic innovation, not the intention of the author. Insofar as the Cambridge School focuses on the contextually bound authorial intention rather than the reception of a text by a contextually bound audience, it is not in fact describing the linguistic innovation that justifies the analysis of the classical texts in the first place.

If we accept that linguistic innovation is a perlocutionary act, the status of the classic texts is once more obscured. We might ask again what the point of studying the classic texts is, if they are only the cause and not explicitly the effect of linguistic innovation. We want to know the effect past events have had on our contemporary language; the causes of such events, like the intentions of its actors, are largely irrelevant.

In a later article, Skinner partly acknowledges this and backtracks from what he calls his earlier, “polemical” work.Footnote ²¹ He instead claims the classic texts ought to be treated merely as data for the explanation of historical events, with his contextualism reclassified as sociology rather than philosophy.Footnote ²² This new stance is even more subject to the dangers of relativism for which the externalist causal account of meaning can provide an antidote.

2. The Causal Theory of Meaning

Our account is a defense of the early, “polemical” Cambridge School project. It is motivated by the view that Skinner's interpretation of meaning change is problematic and Kripke's causal theory of meaning is the fix. The philosophy of language has developed since Wittgenstein, Grice, and Austin. The Kripkean revolution has swept through analytic philosophy departments worldwide,Footnote ²³ following the publication of Naming and Necessity in 1972. Variations of Kripke's causal theory of meaning have garnered near-universal consensus among philosophers of language.Footnote ²⁴ We argue that the causal theory of meaning, despite being entirely ignored by historians of political thought, is ideally suited to justifying the Cambridge School's focus on linguistic innovation.

The causal theory of meaning demonstrates how the context of the past is important for the contemporary meaning of terms. The classics of political philosophy are more likely to be responsible for fixing the reference of these terms than any other source. This justifies their study because this reference is preserved in political terminology, even when it is no longer epistemically grasped. The meaning of the term can thereby entail both existential and normative assumptions that the speaker now believes to be false but that were salient to the original reference-fixing context of the term. This is the lesson of the causal theory of meaning. It tells us that (1) we accommodate this conceptual baggage unknowingly and (2) this ignorance is an ignorance of context, specifically an object or relation in the historical context that was fixed to the term as its reference.

The causal role of the text here is part of its perlocutionary force—the force of its performance—not its illocutionary force, and is therefore better described by focusing on the text's audience rather than its author. A term's meaning shifts if and only if an audience accepts the shift and incorporates it into subsequent discourse.

The idea behind Kripke's theory is that we fix types and token objects we observe in our own context to terms in our language. We observe things, patterns, and relations in our own context and then fashion terms and markers (that is, words) to refer to these things when communicating information from one context (say, our personal context) to another (say, the personal contexts of our audience). There will be significant overlap of referents between the personal contexts of people living in the same time and place and less overlap between people living continents and centuries apart.

To the extent that we are mistaken about the things we refer to that are particular to our own circumstances, we will be mistaken about them when we think about them in other circumstances as well. Kripke draws this out by focusing on nonactual possibilities. We will stipulate these possibilities by importing the objects and relations existing in the actual state of affairs that we baptized with names.

It is perhaps not clear yet how we come to grasp a name semantically if it is not to associate a certain set of descriptions with it. Kripke's following sketch has proved influential:

He also describes it in his supplementary lecture Reference and Existence:

This is the causal theory of meaning. The subject of the baptismal event that starts the causal chain is the individual i itself. The individual i here stands thereafter as the meaning of the name “Feynman.” The meaning of “Feynman” exists in the external, material world, which is why Kripke's semantics is called “semantic externalism.” The meaning of “Feynman” is not some idea internal to the minds of those competently using the name: it is the actual person Richard Feynman. A term is passed down through a linguistic community, where each use of the term is causally linked to the initial baptism, and through this baptism to the term's external reference. As Putnam puts it, “Cut the pie any way you like, ‘meanings’ just ain't in the head!”Footnote ²⁷

Instead the meanings of our political terms are often bound up in the actual reaction of a usually long-dead audience to certain texts. These reactions are often unconsciously entailed by competent language-users in argument. A 1949 panel of the American Political Science Association concluded that the classics are still important because they made us “aware of our own mind.”Footnote ²⁸ The causal theory of meaning gives credence to this mysticism. It suggests the meaning of our ideas is not simply internal but also external: external objects and relations are partly constitutive of our internal thoughts. The conceptual baggage arises from the aspect of the external world anchoring the meaning of the political term. We might be mistaken about this aspect of the external world and this might lead us to unwittingly accommodate bias in our political terminology.

Linda Zerilli, for instance, accuses Jean Elshtain of such bias in using a predetermined language in her analysis of Sophocles's Antigone, translating “the foreign, dissonant voice of Antigone into the more familiar, reassuring voice of mothers and/or citizens … situating themselves in relation to the accepted terms of debate (public/private).”Footnote ²⁹ Zerilli argues women have to “cross dress” talking in a language filled with “masculine images.” Even when a seemingly neutral concept like the public/private distinction is used, unbeknownst to the users they might be accommodating bias. In particular, the referent of “private” is quite plausibly fixed to the more traditional family and the inequalities it entails. This echoes Kripke's suggestion that stipulations of states of affairs will import objects and relations in the context of utterance into the circumstances the utterance is evaluating.Footnote ³⁰ To the extent that we are mistaken about the object or relation in the context of utterance, we will be mistaken about it in the circumstances of evaluation.

When thinking about states of affairs, our own minds, even when we are thinking about states of affairs long gone or never even realized, will always import references from our current context into our stipulations of these states of affairs. When we think about Feynman, our thoughts concern the referent that was initially baptized “Feynman” in the external world by our linguistic community, whatever the characteristics of that individual turn out to be. Likewise, when we think about “the state” it is the referents that were initially baptized by the term, whatever the characteristics of those referents turn out to be. Objects and relations constituting the context of a reference-fixing event often stick to the term for subsequent usage; just as the individual i who was baptized “Feynman” sticks to the name Feynman in subsequent utterances of the name.

We should not, however, take it on trust that the classics changed the reference of our political terminology. Consequently, we should not take it on trust that the classical canon is a worthy object of study to help us be more “aware of our own minds.” Blurring illocutionary and perlocutionary force has caused the importance of the classics to fade from scrutiny within the Cambridge School. It has been taken largely on trust, despite Skinner's earlier challenge to the orthodox canon (arguing for its expansion to pad out the intellectual context of various eras). It is also despite a growing concern for the “cultural insularity” of such methods.Footnote ³¹ As Arlene Saxonhouse writes, “while detailed summaries of the content of minor works are a large part of the study, these minor texts continue to circulate around just those major authors familiar to those who study the canon. … The texts may begin to skin into the landscape of discourse, but they never disappear.”Footnote ³²

We counter that any discursive event that shifts a term's reference should be the subject of the history of political thought. While we will show that the ideal constructions of philosophers make them more likely reference-fixers (see section 7), the relevant text could just as easily be a holy book, newspaper article, court case, or even a parliamentary speech documented in Hansard (think of the political speeches of Sojourner Truth, Edmund Burke, and J. S. Mill, for instance). We should not prejudge the text's perlocutionary force. This is the first step in our two-step methodology. This focus on reception—specifically by the linguistic community that generated our contemporary idioms—might exclude certain texts previously included within the canon. It is a stretch to say we are part of the same linguistic community as the ancient Greeks, for example, with the implication that it would be more valuable to study the ancient Greeks by way of Renaissance thinkers’ interpretation of them. When we focus on the classics’ perlocutionary causal role in reference-fixing a large chunk of our political vocabulary, their contemporary relevance is clear. We read the classics to discover the outdated passages, strange arguments, and obvious posturing, that led audiences to accept new terms and revise old ones. Kripke's causal theory of meaning strongly suggests we unknowingly accommodate some of these anachronisms within modern political terminology. Catching these anachronisms is the first step towards debugging them in political argument.

3. An Example

Our picture perhaps differs from the interpretation Cambridge School practitioners would give to their own work, yet it solidifies their work's philosophical importance. For one, it stresses the philosophical relevance of Skinner's own etymological investigation of the term “the state,” and the burgeoning historical research program that has followed.Footnote ³³ Skinner notes the evolution of the term from the sixteenth century where it referred to the princely status of individual rulers, to its seventeenth-century use referring to a permanent legal construction. With the aid of older theories of meaning, he contends that we better grasp the concept of the state by describing the evolution of the term “the state.” Skinner sees change in meaning as a gradual process, a bit like a game of Chinese whispers, where shifting descriptions associated with the term imply that the meaning of the term itself is changing.

Yet Kripke's causal theory of meaning suggests not every descriptive mutation of the reference of “the state” is important to understanding its meaning. There are many incidental descriptions that are irrelevant to its meaning. While Skinner notes a number of important nodes in the history of “the state” by highlighting a historical series of redescriptions of the term's extension, his selection could be construed as arbitrary. Redescriptions of the term's reference do not necessarily imply that the term has changed its reference, which means new descriptions of the term do not imply, to follow Kripkean semantics, that it has changed meaning.

Skinner suggests it was with Hobbes's Leviathan that “the state” mutated into its modern guise.Footnote ³⁴ Hobbes used a metaphor to fix “the state” with its modern meaning: the state is the metaphorical seat upon which the sovereign sits. This was a decisive shift in the term's reference. “The state” no longer referred to a fleeting status of princes but to a stable legal construction that existed over and above the passing fortunes of individual politicians. This is a perspicuous case of reference-shift, which Kripke implies is the only way terms described by his causal theory of meaning can change their meaning. Simple redescription does not have the same effect. Kripke paraphrases Gareth Evans's example of “Madagascar” to demonstrate this kind of meaning change:

With Hobbes's appropriation of “the state” the reference of the term was not preserved, thus starting a new historical chain of transmission, much like Marco Polo's appropriation of “Madagascar.” This new chain is arguably still the same chain fused to our modern usage of “the state.” Skinner does much to explain Hobbes's importance in this respect, which is to say the reception of Hobbes's Leviathan serves as the reference-fixing event for our modern term “the state.” The causal theory of meaning takes a change in meaning of a term to be a complete and decisive baptism of a new referent.

At no point in Leviathan does Hobbes give a nonmetaphorical definition of what the state is. It is always by way of the metaphor of the seat upon which the sovereign sits. The sovereign is described in similarly abstract fashion: “an Artificiall Man” who is created so that “A Multitude of men, are made One Person, when they are by one man, or one Person, Represented.”Footnote ³⁶ We can take such descriptions as demonstratives pointing to the intended object to be fixed as the reference of “the state.” These descriptions are not themselves definitions, far less necessary and sufficient conditions. “The state” does not literally refer to a seat or a throne; it is a social kind, not a physical object.

In Kripkean fashion, Hobbes immediately follows his metaphorical demonstrative with a number of paradigm samples: treating “commonwealth” as synonymous with “state,” he asserts that “all Christian kings, popes, clergy, and all other Christian men make but one Commonwealth: for it is evident that France is one Commonwealth, Spain another, and Venice a third, etc.”Footnote ³⁷ “The state” directly refers to the social kind Hobbes was referring to in his seventeenth-century context with sample demonstrations. This arguably reveals the unconscious pull of context in our modern political arguments over the limits of the state.

“The state” had its reference fixed with seventeenth-century objects and kinds. Europe then was teeming with clear and distinct sovereigns—kings and queens—who were undisputedly at the apex of power in their jurisdictions; in the twenty-first century, the encroachment of parliaments on the powers of the monarchy has obscured the power hierarchy. The rise of transnational corporations means the prospect of capital flight constrains the policy of would-be sovereigns, further obscuring their social position. With this in mind, we suggest Kripke's semantic position implies “the state” might not have a reference in the modern world, despite our frequent use of the term as if it did.Footnote ³⁸ We study the history of political thought in this respect to determine whether “the state” is now a dead metaphor, whose deathFootnote ³⁹ could be responsible for considerable swaths of humbug when contemporary argument turns to the “limits and duties of the state.”

The question Skinner's work begs, according to our picture, is whether Hobbes's audience did in fact accommodate this mutation of “the state” into common discourse and what subsequent mutations entail. If “the state” is no longer linked (if it ever was) to a historical chain of transmission back to Hobbes's Leviathan, this is all the more reason to take etymological history, as championed by the Cambridge School, seriously.Footnote ⁴⁰

4. Applicability to Appraisive Terms

Applying the causal theory of meaning to social kinds like “the state” and “political party” is controversial, given that the typical examples of semantic externalism are either personal pronouns like “Richard Feynman” or natural-kind terms like “gold” and “water.” Even more controversial is applying it to appraisive terms such as “liberty,” “justice,” and “rights.” Yet we maintain there are important similarities such that the causal theory of meaning can be used to describe the meaning of these terms.

The meanings of terms like “Richard Feynman,” “gold,” and “tiger” are all commonly considered explicable with the causal theory of meaning.Footnote ⁴¹ The testFootnote ⁴² is that the referent can be permissibly described in multiple and sometimes contradictory ways by semantically competent speakers.Footnote ⁴³ Speakers do not need to have a particular description in mind to meaningfully refer with the term. If it is correct to treat differing and contradictory descriptions as more than mere verbal disputes, the terms’ meanings must be causally anchored in an external reference. A chemist might have inherited the term “gold” from their parents and use it with the intention of referring—though in a more precise fashion—to what their parents refer to by the term, as they intended with their parents, and so on. At some stage in the chain of inheritance an external paradigm sample of gold would have had to have been baptized “gold.”

The trade-off to this externalist presupposition is that we can mistakenly apply the term despite using it in a semantically competent fashion. In the same way we can be mistaken that a metal is gold owing to our ignorance of gold's atomic structure, we can be mistaken that the things we call “states” today are states, owing to our ignorance of the historical context in which the reference of “state” was baptized. The chemist can correct their parents’ use of the term by pointing out that their parents’ wedding bands, surprisingly, are not in fact gold (on analysis the chemist realized they are gold-plated silver). Despite associating different descriptions with “gold”—the chemist knows gold is the metal with atomic number 79 whereas their parents do not—the chemist and the the chemist's parents are not engaged in a verbal dispute. They are all meaning to refer to the same thing.

Likewise, when somebody hears the word “liberty” for the first time, they subsequently use it with the intention of referring to the same thing as those they inherited the term from. They might apply the term to different countries than their predecessors did, but this does not change the fact that they intend to refer to the same essential thing as their predecessors. While Dworkin is right to note political terms are different from natural-kind terms because “liberty has no DNA,” he is just as right to note that political terms refer in the same way as “gold” and “tiger” in the sense that speakers treat them as having a core or essence: “The physicist helps us see the essence of water; the philosopher helps us see the essence of liberty.”Footnote ⁴⁴ The intention language users have to refer to essentially the same thing as each other makes political terms explicable with semantic externalism. We will show that it does not matter that one is a physical essence and the other is normative, in the same way it does not matter whether they are fictional or real.

Different ideological viewpoints describe freedom differently. Those on the left might identify it as a social position an individual is in when they fully realize their self-interest while those on the right might identify it as the property of noninterference. The two ideological positions will therefore likely identify different things as “free”; they will use the term “freedom” differently, just as the chemist uses the term “gold” differently from their parents. Yet participants in ideological disputes are not prone to admitting their debate is purely verbal.Footnote ⁴⁵ They are of the opinion the other side has a wrong conception of freedom owing to misinformation, not that the other side just means something different with their use of the term “freedom.” Participants tend to proceed as if the dispute were due to informational asymmetries.

What is passed on when speakers inherit a term is the term's reference, not a description of semantic content. This reference is the term's meaning. In the case of appraisive political terms the reference is, as we will see, an object or relation fixed in a nonactual ideal. There are no descriptions of a term's reference that we necessarily pass down when we pass on the term's usage through a linguistic community; we just pass on the reference, whether speakers are aware of that reference's nature or not.

The arguments Locke employed in Two Treatises of Government, for instance, spill over into the reference of “property” roughly in the way John Dunn describes in his book The Political Thought of John Locke. Dunn's book is a classic of the Cambridge School. He sets himself the task of demonstrating that a particular form of theology is “a necessary condition for the cogency of many of [Locke's] arguments and that there is every reason to believe that Locke himself would have assented to this judgement. If, then, Locke is not judged to have possessed a defensible theology, it is hardly remarkable … that the residue of his thought should provide no coherent account.”Footnote ⁴⁶ Dunn draws out Locke's reliance on the distinctly Protestant “calling as that station in life to which it had pleased God to call a man.”Footnote ⁴⁷ It is up to each individual to work out their distinct calling for themselves and then to industriously labor to meet it. The Protestant work ethic underpins Locke's argument for his distinctive reference-fixing of “property,” which is arguably the reference that has stuck to the term into our contemporary linguistic community.Footnote ⁴⁸ This is despite modern political argument over taxation usually being intended in a strictly secular sense. This disjunction between intention and reality plausibly suggests why such argument so quickly breaks down.

Locke constructs an appraisive ideal in order to evaluate political circumstances: a cosmic order created by God and working toward his will. Locke thinks God “created every part of it for his own purposes and he created each part of it with a defined relationship to the purpose of the whole … an ordered hierarchy … in which every species has its station, its rank.”Footnote ⁴⁹ Each individual should both apprehend the will of God for their particular station in life and then follow it.

Dunn argues that this theological ideal was common ground between Locke and his audience within Locke's seventeenth-century context. Dunn suggests there is enough textual support to believe Locke was aware of this common ground and used it to shift the reference of “property” as used in Patriarcha, the Tory defense of executive authority by Robert Filmer, to a reference that was natural or prepolitical.

Locke used his infamous mixing thesis as an analytical cog to shift the reference. He argued that it is correct to apply “property” to ourselves by virtue of God's calling, not by virtue of any title deed granted to us by authority. Locke premises his mixing thesis by drawing on the theological ideal that he took as common ground between him and his audience. Given we have ownership over ourselves by virtue of our calling from God, when we transform goods with our labor, we mix part of ourselves with that good. If we till some soil, we mix part of ourselves with the land, and therefore own it. Were somebody to subsequently interfere with the land, it would therefore be tantamount to interfering with one's person and, with it, God's calling.

There is something unquestionably bizarre about the mixing thesis from a contemporary perspective. It is analytically incoherent in the abstract. If one owned some orange juice and then mixed it with the ocean, would this imply one owned the ocean? Locke's proviso that people leave enough for others places checks on this analytical absurdity. Robert Nozick, who quite explicitly reworks Locke's premises in a late twentieth-century context, likewise suggests appropriation of property is legitimate only so long as nobody is made worse off by its initial appropriation from the commons.Footnote ⁵⁰ The question that is left over is, Made worse off at what point? At the initial point of appropriation? In the present? In the future? When?

The Lockean answer is simple: illegitimate appropriation amounts to appropriation barring anybody (present or future) from meeting God's calling in their own way. There is arguably no definitive analytical response independent of this theological ideal; we are duped into intuitively thinking there is by virtue of our ignorance of its necessary theological implications. Locke's theology has spilled into the reference of “property,” embedding itself in subsequent usage of the term without speakers generally being aware of it.

5. Spillage from Ideal Constructions

So, Locke's theological assumptions have spilled into the reference of the term “property” by virtue of the context in which “property” was fixed with its modern reference. This is unorthodox in the philosophy of language: it is well established that the argumentative context involved in the baptism of Feynman and gold does not spill over into the reference fixed to “Feynman” and “gold.”

The difference between Feynman and gold, and property and liberty is that the former have actual paradigms whereas the latter have nonactual paradigms. Locke's theological ideal is an ideal where everybody meets God's calling in their own special way, a state of affairs Locke thought did not apply in actuality. This is the crucial difference between triggering reference spillage and not.

The clearest case of reference spillage is with the kind of directly referential name that most obviously does not have an actual paradigm: that is, a term that refers to a fiction (for short, a fictional name). For much of Naming and Necessity’s lifespan, the status of fictional names in Kripke's causal theory of meaning has been obscure. In 2013, however, Kripke published Reference and Existence to illuminate this obscurity. His distinctly nonpolitical example of Sherlock Holmes shows how, first, nonactual possibilities entail reference spillage and, second, how the name is semantically equivalent to appraisive political terms.

If we inherited the name “Sherlock Holmes” from the causal chain leading back to an audience initially reading Conan Doyle's novels, we would simply be wrong if we thought “Sherlock Holmes” referred to a dog. The name refers to the famous consulting detective. What if Sherlock Holmes had never become a famous detective? He would not be the same man referred to in Conan Doyle's fictional world. Certain features audiences accepted about Sherlock Holmes in the books spill into the reference of Sherlock Holmes.Footnote ⁵¹ Ignorance of the novels might mean we now think Sherlock Holmes lived on Gloucester Place, not Baker Street. If we are talking about an individual i who never lived on Baker Street, then we are not talking about the same Sherlock Holmes. While Sherlock in the fictional circumstances of evaluation obviously has the capacity to live elsewhere, it is nevertheless still necessary that Sherlock lived on Baker Street from our nonfictional context of utterance.Footnote ⁵² This is because the name “Sherlock” is indexed only to the object in the particular fictional worlds in which Holmes did in fact live on Baker Street, meaning Holmes lived on Baker Street in every world in which he exists. If a description is true of an object in every world in which it exists, then it is necessarily true of the object.

Sherlock Holmes in the BBC adaptation of the story set in contemporary London might be vaguely similar to the Sherlock Holmes of Conan Doyle's world (call him Sherlock Holmes_BBC), but he is not the same Sherlock Holmes (call Conan Doyle's character Sherlock Holmes_CD). The characteristic of living in nineteenth-century London is a necessary characteristic of Sherlock Holmes_CD.

These necessary characteristics attach to the meaning of the term “Sherlock Holmes” irrespective of whether language users know it. When we talk about Holmes, for instance, we accommodate a fictional proposition about a man who lived in the nineteenth century irrespective of whether we are aware of it. It is the conceptual baggage accommodated with the use of the name “Sherlock Holmes” by virtue of a past audience passing down the name of the hero of their interpretation of Conan Doyle's story. Properties that would be taken as contingent for nonfictional names are turned into necessary properties when they are fixed to the reference of fictional names. This is what we mean by reference-spillage. When there is no actual paradigm we can point to in the actual state of affairs, many features of the reference-fixing context inadvertently spill into the reference of directly referential terms. This provides the model for appraisive terms that are likewise fixed with nonactual possibilities.

6. Reference-Fixing for Appraisive Terms

Assertions using fictional names are constitutive of fictional assertions.Footnote ⁵³ Appraisive political terms are similarly constitutive of appraisive assertions. The claims “Australia is a regime that protects the fundamental freedom of its citizens” and “They took the baron's property” are appraisive. The first appraises Australia positively by virtue of the term “freedom.” The second appraises the subject's actions negatively by virtue of the term “property.” Likewise, the claim “Sherlock Holmes lives at 221B Baker Street” is fictional. It is a fictional assertion by virtue of the inclusion of the term “Sherlock Holmes.” If the name referred to the nonfictional entity “Saul Kripke,” the claim would not be fictional, it would be nonfiction and false.

The paradigm of appraisive political terms like “justice” is fixed by a description and justification of a nonactual ideal, just like “Sherlock Holmes” was fixed by Conan Doyle's description of a nonactual fictional world. In The Republic, Plato, through Socrates, looks to identify “justice” by first reconstructing the “ideal man” and then the “ideal city.” He fixes the reference of “justice” after a long exposition and justification of his ideal city by fixing a particular structure within the city to the term. Socrates says, “The next step is to look inside [the ideal city]. … Then perhaps we shall find some way of seeing just where in the city justice is.”Footnote ⁵⁴ Where there was much attention given to the justification of Socrates's ideal republic, no explicit justification is given to this reference-fixing act. Socrates concedes he is at a loss, and asks Glaucon to pray with him from “the shadows” to fix “justice” to a referent.Footnote ⁵⁵ He suddenly has a revelatory “Aha!” moment and fixes “justice” to the state where citizens of the ideal city are in the occupation best suited to their nature.

This somewhat arbitrary process, we suggest, is the archetype for how appraisive political terminology is fixed. The substantive argument is almost always directed towards motivating the appraisiveness of the ideal, not the decision to fix something within the ideal to a term. When we use appraisive political terms, we are usually aware that the paradigm is not actually manifest, despite it being something we are looking to approximate. Socrates claims the paradigm of justice is nonactual by virtue of his ideal city being nonactual. He asks, “Are we then going to decide that the just man must be in no way different from justice itself, but in every way like justice? Or will we be content if he comes as close to it as possible?” He claims the paradigm of justice in the ideal city is “a model,” which can appraise other cities by virtue of how close they approximate it. They will never themselves be the paradigm: “It wasn't our aim to demonstrate that these things were possible.”Footnote ⁵⁶

This interaction of the nonactual ideal with reality is analogous to directly referential fictional terms. A fan of Sherlock Holmes might attempt to imitate Holmes's model as best they can by starting out as a consulting detective. They are not imitating the broad class “consulting detective,” they are imitating Sherlock Holmes. They are well aware that they will never themselves be Holmes, but they can certainly be more like him than anybody else. It is not incoherent to say they vaguely resemble Holmes, despite the former being real and the latter fictional.

As the fan might be mistaken about certain details of features necessary for being Conan Doyle's Sherlock—they might be unaware of Sherlock's opium addiction, having not yet read the earlier novels—a converser can be mistaken about the details concerning the nonactual paradigms of appraisive political terms. Few individuals in a linguistic community have read the texts that fixed the reference of their political terminology, let alone investigated their historical context.

Collingwood claims, “The Republic of Plato is an account, not of the unchanging ideal of political life, but of the Greek ideal.”Footnote ⁵⁷ What is more, the renaissance writers who rediscovered the ancient Greek texts might have only reintroduced “justice” along Platonic lines because of the apparent similarity between the ideal city and a well-ordered monarchy. If this is historically true, and if we have indeed inherited the meaning of “justice” from these thinkers, anachronistic Greek assumptions and anachronistic Renaissance platitudes might explain our intuitions concerning justice. We all know there is something praiseworthy about justice and we all have an idea of its rough general shape, but the reasons we have such praise-giving intuitions is unclear. We are epistemically ignorant of the facts.

This empirical ignorance—and history's role in rectifying it—is demonstrable with a similar make-believe story to Kripke's story of Feynman's baptism. Let us say the Platonic ideal city was accepted by the European intelligentsia of the Renaissance by virtue of its closeness and seeming justification of the divine right of kings. The suggestion of fixing “justice” to a particular relationship in this ideal consequently had strong perlocutionary force and was ultimately accommodated. The Renaissance thinkers passed on their ideas to political pamphleteers who incorporated this reference of “justice” into their prevailing political rhetoric. Those favorable to the king imbibed this appraisive usage of “justice,” bestowing it to the next generation (either as parents teaching their children politics or as part of the school curriculum), who later passed the term down to their children, and they to the next generation, and so on and so forth down through the linguistic community to the present day. With increasing age and increasing acceptance the appraisiveness of its reference became naturalized as common intuition, all as the historically contingent specifics of the Platonic ideal's justification fell out of mind. The contextual beliefs that fostered the acceptance of Plato's ideal city, Locke's Protestant “calling,” Kant's perpetual peace, Hobbes's state of nature (inversely), and Rousseau's natural state are accommodated in the terminology these ideals were used to fix. If this is the same terminology we use today, and if such beliefs are now anachronistic, we should question our intuitions.