What exactly are Gadamer's grounds for denying the existence of a uniquely correct interpretation of a text, object, or event? I begin by showing the inadequacy of two arguments for his position. I then turn to a third more promising argument that objectivity is not possible because the object of understanding is not determinate, but rather constituted anew by each act of understanding. My goal in this paper is to provide a fuller justification for the third argument and thereby defend Gadamer's position. I do so by reformulating this third argument in terms of relational properties so as to establish that the knower's situatedness plays, as Gadamer himself insists, a positive, constitutive role in the process of understanding. A major advantage of this account is that it offers an explanation of how pluralism can recognize criteria for determining correct interpretations and thereby avoid a pernicious relativism.

The first argument in support of interpretive pluralism is rather simple. Objectivity is not possible because an inquirer's prejudices or prejudgments (Vorurteile) are ultimately inescapable (see e.g. G260f.; E 276). The inescapability premise is one that many philosophers would regard as true. It does not, however, entail Gadamer's conclusion because even if objectivity in the form of a total break with historically specific precommitments is an impossibility, one might still hold that it is a suitable regulative ideal for understanding, i.e. an ideal that permits not realization, but at least approximation. So the impossibility of overcoming historical situatedness does not itself entail that objectivity does not or cannot serve as an ideal. It fails to establish Gadamer's anti-objectivism.

The second argument emerges from passages in Gadamer that stress the Diltheyan idea that our historical prejudices connect us to and make possible our understanding of the object in question. On this view, objectivity is not possible because understanding requires translatability and translatability requires a shared background of meaningfulness (or in Gadamer's language: prejudgments and tradition). This argument is also too weak to establish Gadamer's conclusion. An objectivist could concede that a common background of meaningfulness may be necessary for making sense of an object of understanding, especially for grasping all its nuances. But the objectivist could go on to argue that this common background of meaningfulness is a heuristic device that must eventually be isolated and subjected to impartial scrutiny. In other words, the process of understanding human phenomena might be thought to have two stages. The first stage consists in sharing or appropriating a set of background prejudgments. The second stage consists in taking distance or working oneself free from the operative precommitments and, in general, approximating the ideal of an unbiased, objective stance towards those very precommitments. It is the superimposition of distanced impartiality onto a shared background of meaningfulness that makes for a sensitive yet balanced understanding and that provides the ideal for the one correct interpretation for which inquirers strive. So the need for a shared background understanding is compatible with objectivism and the denial of interpretive pluralism. It too fails to give Gadamer what he wants.

I turn now to what I consider to be the best argument for Gadamer's position. According to this argument, the inquirer's historical situatedness plays a constitutive role in the act of understanding because without it there would be no complete, fully determined object to understand. This idea is suggested by passages in Truth and Method that deny the existence of any sort of "object in itself" underlying inquiry in the human sciences (G 269; E 284f.) or refer to the historical object as a kind of "phantom" (G 283; E 299). Here objectivity is not a possible ideal because the object of understanding is fundamentally underdetermined; it is constituted in part by the horizon of the specific historically situated knower and changes according to what that horizon is. If true, this third argument, unlike the first two, does in fact entail interpretive pluralism since it denies that there is any one unchanging object to be understood. But is it true and on what possible grounds?

Gadamer is not explicit about his support for the incompleteness claim. Yet consider the following passage from his discussion of the importance of temporal distance:

Gadamer is saying that the object of understanding is incomplete because it, or its "meaning," is revealed differently as a result of subsequent events that brings about different points of view. Consider, an artwork such as a Cubist painting by Picasso or Braque, a text such as the American Constitution, or a historical event such as the Russian Revolution. Our understanding of these "objects" is quite different in virtue of the temporal distance that separates us from them. The importance of temporal distance here consists not in any alleged growth in impartiality, but in the way in which more recent events have brought out new aspects of or "retrodetermined" the earlier phenomena. In the case of cubism, there is the subsequent development of increasingly abstract painting. In the case of the American Constitution, there is the two-hundred year history of new issues and cases and a continuing tradition of judicial interpretation concerning the Constitution's original provisions. In the case of the Russian revolution, there is the occurrence of Stalinist totalitarianism, eventual economic stagnation and finally the collapse of Soviet Communism. The point is that the Cubist paintings, the Constitution and the Russian Revolution not only appear in a very different light, but have come to have different relational properties as a result. They have become phenomena that bear certain new (causal and non-causal) relations to objects and events that came after them. It is in this sense that the object of understanding can never be completely grasped. In Gadamer's terms, the object itself is "constantly being formed" ("in beständiger Bildung begriffen") (G 277; E 293). Now if the object or event grows or changes over time, this means that there is no single, enduring correct or objective understanding of it.

Let me turn now to two possible objections to my reconstruction of Gadamer's interpretive pluralism. First of all, a skeptic might contend that relational properties are not ontological properties of the object at all, but only epistemological items that merely introduce changes in the ways we describe an ontologically determinate object. On this view, when a later historical event leads us to see an earlier historical event differently, it is only our description of the earlier event that changes, not the earlier event itself. This position which denies the ontological reality of relational properties is mistaken for the following reason. It is true that our descriptions of earlier events change as a result of later events. Yet it is not just our descriptions that change. Relational properties are not features of our descriptive predilections, but of the events themselves. Our descriptions sometimes change because we have changed, but they sometimes change because the objects relational properties have changed. For example, if a person describes the Russian Revolution differently because she has undergone a political conversion, this descriptive change is a result of a change in that person's epistemic or attitudinal makeup, not in the event itself. If, however, a person describes the Russian Revolution differently because the Revolution has come to bear new relations to new events, then it is not the person that has changed but the Revolution, insofar as it now has new relational properties (e.g. the property of having led to a 70-year failed alternative to capitalism). For this reason, relational properties must be regarded as ontologically real; though they may lead to new descriptions, they are not merely changes in the epistemic makeup or descriptive activities of persons.

Two features of Hirsch's theory should be noted here. First, unlike Gadamer, Hirsch is concerned not with understanding in general, but only with understanding literary texts. We can set this difference aside. Second, Hirsch interprets the stable meaning of a text to consist in the author's intention. It is important to point out that the stable meaning of a text might be construed differently: not in terms of authorial intention, but in terms of the intrinsic properties of the text — a view sometimes called formalism. So the idea that a text's meaning is fixed might be defended in two ways: either by identifying a text's meaning with the author's intention or with its formal, intrinsic properties. In either case, however, if Hirsch is right about the basic distinction, then Gadamer's anti-objectivist, interpretive pluralism goes down with its conflation of the interpretation of fixed meaning and criticism's interest in shifting significance.

Now how might one defend Gadamer (and my reconstruction of his position) against Hirsch's point about the difference between shifting relational properties and fixed meaning?

My response is this: Although the distinction between relational and intrinsic is correct and essential to making sense of Gadamer, Hirsch is wrong to think that the object of understanding  — the object that we seek to understand and eventually do understand (when our efforts are successful) — is the object shorn of all its relational properties. In other words, significance or relational properties are always operative in and constitutive of our encounter with that which we seek to understand. What is the object of understanding or interpretation? Is it, should it or can it ever be stripped of its relational properties? Or are these relational properties integral to it?

Let us begin with a (single-authored) text. Against Hirsch's view, the following points can be made. First, as anti-individualist theories of mind have shown, the identification of something like an intention will depend on certain facts about the environmental context in which that intention is situated. So intentions already involve context. Second, and more important, it seems altogether odd to think that the meaning and uniqueness of a work is identical with and exhausted by the author's intention. There are (at least) two reasons for thinking that there is more to the text than what the author intends. First, as Gadamer writes: "What expression expresses is not merely what is supposed to be expressed in it — what is meant by it — but primarily what is also expressed by the words without its being intended — i.e., what the expression, as it were, betrays." (G 318; E 335f.). In other words, there can be much more in an utterance or expression than what a person had in mind. Individuals are not always the best judges of their own verbal behavior. The second reason concerns the reader not the author. Why do we try to understand what we try to understand? Why do we read Shakespeare or Max Weber? Is it really only in order to reconstruct their psychology, their thought or their will? Or is it not much more a matter of trying to understand the subject-matter that they address? In most cases a wish to reconstruct intention remains ancillary to understanding the subject-matter. Gadamer makes this point repeatedly. He states that understanding is always a matter of "coming to an understanding about something" (G 168; E 180; emphasis added) and that "the hermeneutic task automatically turns into a problematic about the subject-matter (eine sachliche Fragestellung)" (G 253; E 269). While it may be interesting and possible, to some extent, to reconstruct an author's intention, the object of understanding is not limited to antiquarian interests. The text always exceeds the author's designs.

The formalist view falls victim to similar difficulties. In order to identify a text's formal semantic properties, social and linguistic context must be brought in. So intrinsic properties are never really wholly intrinsic in the first place. Second, our interests typically concern more than just a text's intrinsic properties. Even if it is possible, to some extent, to perform a kind of phenomenological reduction by bracketing out the relations the text bears to other things in order to focus exclusively on its intrinsic properties, to do so is to engage in an activity quite different with the more common and more natural ways in which we understand. What we usually understand (or strive to understand) when we understand the meaning and uniqueness of a text is not the text divorced from but illuminated by its relations to what lies outside of it.

If we turn from texts to historical events, a critic of Gadamer might argue that the historical event consists solely in its intrinsic properties. But the anti-relationalist position seems even weaker here. Restricting our understanding of events (and actions) to intrinsic properties would make it impossible to refer to events in many of the ways that we typically do. Consider the following examples. We could not understand a shooting as a killing if the victim were to die some time after the shooting because the killing involves a relation between the shooting and another event, the subsequent death of the victim. Nor would we be able to understand the bombing of Pearl Harbor as the immediate cause of U.S. military involvement in World War II because this understanding of the bombing involves relating it to later events.

What we understand when we understand are objects in terms of their intrinsic and relational properties. The meaning and uniqueness of the phenomenon of interpretation is always (notwithstanding certain specialized efforts at grasping intentions and supposedly formal properties) bound up with its significance. And because relational properties vary for the reasons discussed above, the object of understanding is never once and for all determined.

Let me conclude by indicating briefly how my reconstruction of Gadamer's interpretive pluralism saves it from an "anything-goes" relativism. First, as I have argued all along, the fixed intrinsic properties constitute one central source for rational constraints on validity in interpretation. As for relational properties, they are not at the whim of the interpreter but depend on the specific conditions under which an object is presented and are intersubjectively verifiable. The fact that the Russian Revolution of 1917 eventually led to the collapse of the Soviet state is a relational property that is there for all to see. The same goes for other relational properties. Whether they obtain is relative to the interpreter's position, but not simply up to the interpreter. So, interpretive pluralism need not be anarchic. Indeed, it is thoroughly consistent with the ideals of impartiality, scrupulousness and fidelity to text and event.