Back to David Vessey's Homepage



Gadamer and the Fusion of Horizons




Hans-Georg Gadamer is often criticized for his account of the fusion of horizons as the ideal resolution of dialogue. I argue that in fact it is an excellent account of the successful resolution of dialogue, but only in light of a proper understanding of what Gadamer means by “horizon” and how then horizons are fused. I do this by showing how Gadamer is drawing on the technical sense of “horizon” found in Edmund Husserl’s and Martin Heidegger’s phenomenologies. In the process I show why a prominent criticism of Gadamer’s account of the fusion of horizons, a criticism presented most forcefully by E. D. Hirsh, is mistaken.



Hans-Georg Gadamer’s account of dialogue is distinctive in three ways: it embraces the possibility of dialogue with texts, it takes dialogue to be the model of language at work, and it considers dialogue successful when there has been a “fusion of horizons.” This last view—that dialogue, when successful, resolves itself in a fusion of horizons—is one of his most controversial claims as it suggests that conversations are only successful if they end in mutual agreement. Even if one were to be charitable and allow resolutions such as “agreeing to disagree,” this is too high of a requirement for the goal of all conversation. A conversation between people that fails to resolve a disagreement, but still leads greater articulation on the part of the interlocutors and perhaps greater sympathy for each other’s views counts as a success, even granting that they entered the conversation seeking agreement. Moreover, agreement for its own sake is not a goal of dialogue; interlocutors could come to agreement solely on of the force of personality or insecurity. Such an “agreement” would violate what should be a central tenant of dialogue, namely that the resolution of the disagreement should be based on the evidence for the positions. For many reasons, then, Gadamer's claim that dialogue seeks resolution in a fusion of horizons is regularly criticized by philosophers. Here I will look closely at Gadamer's view, for I think that the accessibility of the phrase hides what is actually a plausible view worth defending as the proper outcome of dialogue.

            To get at his view, it will be helpful to begin with an objection by E. D. Hirsch.[1] Hirsch thinks Gadamer's discussion of horizons immediately precludes something like a fusion of horizons. Hirsch argues that whatever we want to understand is either within our horizon or beyond our horizon. If it's the latter, then it can't be understood, as that's what it would mean to be beyond our horizon. But since it can't be understood there can’t be a fusion of horizons. If it is within our horizon then there are not two separate horizons to be fused, and no fusion takes place. We should conclude, therefore, that if horizons are limits on understanding, a fusion of horizons is either impossible or unnecessary. Here's how Hirsch puts it:

How can an interpreter fuse two perspectives—his own and that of the text—unless he has somehow appropriated the original perspective and amalgamated it with his own? How can a fusion take place unless the things to be fused are made actual, which is to say, unless the original sense of the text has been understood? Indeed, the fundamental question, which Gadamer has not managed to answer, is simply this: how can it be affirmed that the original sense of a text is beyond our reach and, at the same time, that valid interpretation is possible? … If he were true to his assumption of radical historicity, that which he calls a fusion of historical perspectives could not be affirmed at all. If the interpreter is bound by his own historicity, he cannot break out of it into some halfway house where past and present are merged. … For once it is admitted that the interpreter can adopt a fused perspective different from his own contemporary one, then it is admitted in principle that he can break out of his own perspective. If that is possible, the primary assumption of the theory [of being bound by horizons] is shattered.[2]

Something with a different horizon would have to either fall within our horizon, in which case understanding it doesn't involve a fusion of two horizons, or fall outside our horizon, in which case we can’t understand it.

Granted Hirsch is concerned with what he calls “radical historicity,” the view that we can never understand something except from our historical point of view and while Gadamer doesn’t hold this view, the paradoxes Hirsch is concerned about arise anytime someone posits knowledge of a limit on understanding. It would seem the only way out of these paradoxes would be for Gadamer to argue that horizons aren’t limits after all. Yet, he does seem to say quite clearly that horizons are limits.

Every finite present has its limitations. We define the concept of “situation” by saying that it represents a standpoint that limits the possibility of vision. Hence essential to the concept of a situation is the concept of a “horizon.” The horizon is the range of vision that includes everything that can be seen from a particular standpoint. [3]

What is lost in most interpretations of Gadamer, including Hirsch’s, is the realization that Gadamer “horizon” is not using the term in its everyday sense, but as a technical term; the technical meaning has been part of phenomenology since Edmund Husserl introduced it in 1913. The two meanings aren't unrelated, but the technical sense emphasizes some features of the everyday term and deemphasizes others.

Specifically, a horizon as a limit is downplayed in the technical meaning in favor of a horizon as that which expands, that which we can see beyond with a little effort, and that which points toward something more. Although a horizon marks the limit of sight at any moment, it is not an insurmountable limit. Simply walking a short distance, or going to the top floor of a building can help us see beyond our previous horizon.  In fact, most of us know quite well what lies beyond the horizon simply from past experience. Horizons might function as a limit at a particular time, but there are always also gateways to something beyond; it is the latter that Husserl emphasized in his 1913 Ideas.[4]

Husserl pointed out that although our senses only give us incomplete information about an object, we perceive the object as a whole.  So although when looking at a chair we are only presented with one side of the chair, we perceive a chair, not a chair-side. We are not surprised when we move to see the chair has other sides to it, that it's three dimensional, and so on. In fact we would be quite shocked to find out what we thought was a chair was only a chair-faćade.  Likewise when perceiving a person we can often tell who a person is based on very little sensory information. We can recognize someone from the back of their head; were he or she to turn around to reveal that we were mistaken, this itself is a sign that our perception of the back of the head included more than simply the back of a head. Were it otherwise we wouldn’t have been surprised to find he or she was someone we didn’t expect. So perception always goes beyond what is actually given to the senses. According to Husserl, it's not that our mind is working drawing inferences from the perceptual information we receive such that, for example, we first see a field of color and then our mind organizes the color and concludes it is some object or person. We actually see it as some object or person. The additional content that fills out the object is contained in the perception of the object.

Martin Heidegger gave the helpful example of hearing the whistle of a train. When we hear that sound we hear a train, or a train whistle. It's only though an extreme effort can we hear that sound without hearing that it's a train. The same applies for all sounds, they are heard as the sounds of things, not pure sounds. Even sounds we don’t recognize we perceive as the sound of something; when we don't recognize a sound, it's the thing, the object making the sound that we don’t recognize. The same applies to vision.

            Husserl believed that understanding how our minds fill out our perception of objects would go a long way toward telling us about the nature of these objects. Those aspects of an object that are not directly accessible to our senses, but make it possible to see an object as an object, Husserl's called the object's “horizon.”

What is now perceived, and what is more or less clearly co-present and determinate (or at least somewhat determinate), are penetrated and surrounded by an obscurely intended to horizon of indeterminate actuality. … [An] empty mist of dim indeterminacy is populated with intuited possibilities or likelihoods, and only the “form” of the world, precisely as “the world” is predelineated. Moreover my indeterminate surroundings are infinite, the misty and never fully determinable is necessarily there.  … This horizon, however, is the correlate of the components of undeterminateness essentially attached to experiences of physical things themselves; and those components— again, essentially—leave open possibilities of fulfillment, which are by no means completely undetermined, but are, on the contrary, motivated possibilities predelineated with respect to their essential type.[5]

The horizon is everything we are aware of in the perception of an object above and beyond what is given directly to our senses. It's what is “co-given” in the perception of the object that makes it intelligible to us as an object. When we walk around a house we are not surprised to see it has sides, that it has a back, that it doesn’t elevate off the ground, that tree leaves don’t knock it over when they brush against it, that when we look away and look back it remains the same, and that it didn't just spring into existence immediately before we saw it.  All these things belong to the horizon of the perception of the house. In his 1931 Cartesian Meditations Husserl writes that “Perception has horizons made up of other possibilities of perception, as perceptions we could have, if we actively directed the course of perception otherwise: if, for example, we turned our eyes that way instead of this, or if we were to step forward or to one side, and so forth.”[6] These possibilities are not random possibilities, but belong to the essential nature of the object as spatio-temporal.

Husserl divides the kinds of horizons between “internal,” “external, “and “temporal.”  Internal horizons are horizons that arise from the nature of the object either as an object, such as taking up space, or as the kind of object that it is. When we see someone from the back, we perceive him or her as having a face. It belongs to the internal horizon of a head to have a face. External horizons are horizons established by the relation between the object and it surroundings. When there is a telephone pole partially obstructing our view of a house, we recognize the parts of a house on both sides of the pole as still belonging to the same house and as being located behind the pole. We also perceive the character of interaction between objects in virtue of an external horizon. All relations get their character from the external horizon, including the relation of belonging to one spatio-temporal whole with everything else.

To focus on one object requires a bracketing of the external horizons, just as focusing on the sensory givenness of an object requires bracketing the internal horizon. Maurice Merleau-Ponty puts it particularly well when he writes that

It is necessary to put the surroundings in abeyance the better to see the object, and to lose in the background what one gains in focal figure, because to look at the object is to plunge oneself into it, and because objects form a system in which one [object] cannot show itself without concealing others. More precisely, the inner horizon of an object cannot become an object without the surrounding objects becoming a horizon. …  The horizon, then, is what guarantees the identity of the object throughout the exploration. … The object-horizon structure, or the perspective, is no obstacle to me when I want to see the object: for just as it is the means whereby objects are distinguished from one another, it is the means whereby they are disclosed. [7]

 Merleau-Ponty brings together three themes: objects never appear without horizons, the horizons establish the set of expectations that accompany the perception of an object, and object horizon structures are connected to perspectives.  The inner horizon—those characteristics of the object that shape our expectations of future perceptions of the object—and the outer horizon—those characteristics of the environment that shape our expectation of the interaction between the object and its surroundings—are the two most significant horizons for our purposes of understanding Gadamer’s use of the term. For Husserl, though, the temporal horizon is the most important as all objects appear to us as temporal objects, as extended in time just as they are extended in space. Since the inner horizon is revealed by our normal expectations for future revelations about the object, and the outer horizon is revealed as the way the object interacts with its environment, time is the crucial linking theme. Future revelations and future interactions, as well as past histories that made it what it is and established it where it is, are essentially temporal and thus Husserl concludes that it is the temporal horizon that is the condition for all other horizons.

            Notice what the horizons are and what they are not.  They demarcate what can be directly sensed from what can’t be directly sensed, but they don’t limit the perception of the object to what can be directly sensed. On the contrary, the horizons are what make it possible to perceive more than what is directly sensed. Horizons open up possibilities of the object and in doing so provide it with its character. In the case of sight, the horizon is connected to the perspective on the object, it marks what stands out and what recedes and it is constantly changing as our perspective changes either from our movement or from the object moving. In short, horizons are the conditions that provide the meaning for the object, conditions which need to be made conscious for a proper understanding of the object.  As providing the key to understanding our perceptions they draw us forward inspiring investigation and making confirmation or falsification of our expectations possible. Husserl does not stress the limiting feature of horizons, but instead the features of horizons that provide a meaningful context for our experiences and that guide our inquiries. This is clearly contrary to the everyday emphasis on a horizon as a limit.

In his first mention of horizons in Truth and Method Gadamer makes explicit his debt to Husserl:

Undoubtedly the concept and phenomenon of the horizon is of crucial importance for Husserl’s phenomenological research. With this concept, which we too shall have occasion to use, Husserl is obviously seeking to capture the way all limited intentionality of meaning merges into the fundamental continuity of the whole. A horizon is not a rigid boundary but something that moves with one and invites one to advance further. Thus the horizon intentionality that constitutes the unity of the flow of experience is paralleled by an equally comprehensive horizon intentionality on the objective side. For everything that is given as existent is given in terms of a world and thus the world horizon is given with it.[8]

Gadamer's emphasis is consistent with Husserl's: the horizon is not the limit of meaning, but that which extends meaning from what is directly given to the whole context in which it is given, including a sense of a world. As such, it draws us away from what is immediately given toward the greater context that provides the meaning. And naturally the horizon changes as perception changes. Gadamer preserves those essential elements of Husserl's account, but transforms them in a simple, but significant, way: he applies Husserl's account of horizon to propositions. He writes, “every proposition has its horizon of meaning in that it originates in a question situation.”[9] Where Husserl spoke of horizons as making meaningful our perception of objects, Gadamer speaks of horizons as making propositions meaningful. His focus is linguistic rather than perceptual and thus is intimately connected with dialogue.

In the case of sentences, the horizon will be the set of beliefs that make it possible to understand the sentence. These beliefs will not only be about the subject matter of the sentence, but the context in which the sentence is presented. For example consider Aristotle's claim that the highest end in life is happiness. To understand this claim we need to know something about eudaimonia, we need to know something about Aristotle's other arguments and views, and we need to something about what a highest end to life might look like. The background beliefs form the horizon—the condition for understanding the sentence.

Although the term is now more metaphorical than literal, we can still think of horizons as providing perspective and as marking the limits of what can be seen from a particular point of view. Horizons provide perspective by being the implicit and explicit beliefs that provide the context for understanding a sentence; they establish what is significant for understanding (those in the foreground), and what is insignificant (those in the background). Horizons mark the limits of our understanding as our background beliefs and knowledge affect what sentences we understand and how we understand them. And horizons take on a particular point of view as the beliefs reflect individual differences. For example, because of their philosophical commitments and inclinations, philosophers tend to greet new propositions openly or skeptically, depending on how it fits with their other beliefs. Finally, just like the perceptual horizons, intellectual horizons draw us away from the immediate sentence toward that which makes the sentence meaningful. The sentence means something, and we are immediately drawn into investigating the context of the meaning and drawn into asking why someone might assert that sentence. Gadamer writes,

That a proposition is more than the representation of a given objective content means, above all, that it belongs to the whole of a historical existence and that it is contemporaneous with everything that makes its presence felt in it. When we want to understand sentences that have been handed down to us, we engage in historical reflections, from which it is determined just where and how these sentences are said, what their actual motivational background is and therewith what their actual meaning is. When we want to represent a sentence as such to ourselves we must, therefore, represent its historical horizon.[10]

Just as for Husserl the horizon of the object presents the object as belonging to a physical world, for Gadamer, the horizon of a sentence presents the sentence as belonging to a linguistic, cultural, and historical world.

By focusing on understanding rather than simply perceiving, Gadamer sets the bar higher for the existence of horizons. We may perceive things without understanding what they are; likewise we may read sentences yet be incapable of understanding their meaning. In stark contrast to Husserl, then, Gadamer claims that it is possible to have no horizon at all and that “it requires a special effort to acquire a historical horizon.”[11] To return to the visual metaphor, someone who lacks a horizon lacks the visual markers to place their perception in context.  Lacking depth perception, a person would be unable to judge well the relative significance of objects or facts. Likewise, lacking a sense of historical or cultural context a person would be unable to judge the relative significant of objects or facts. Gadamer writes, “A person who has no horizon does not see far enough and hence overvalues what is nearest to him.”[12] For Husserl, a horizon provides a sense of perspective, a sense of the place of an object in its world. In the case of sentences, a horizon would be a set of beliefs that provide the possibility of understanding the sentence in historical context and with respect to its subject matter.

Gadamer’s shift away from perception towards intelligibility through language reflects one of his criticisms of Husserl, namely that Husserl prioritizes too highly sense perception in the acquisition of knowledge, and, in general, in explaining a person’s fundamental relation to his or her environment.[13] It also reflects Heidegger’s use of the term leading up to Being and Time. Heidegger used the word sporadically prior to Being and Time where he introduced the idea that “time as the possible horizon for any understanding of being.”[14] Speaking explicitly of temporal horizons, Heidegger says that the horizonal unity (of past, present and future horizonal schemas) “makes possible the primordial connection of the relations of the in-order-to with the for-the-sake-of which;”[15] that is to say, the horizon makes possible understanding something as having relevance within a world. In The Basic Problems of Phenomenology he defines the horizon as “that towards which each ecstasis is intrinsically open in a specific way…the open expanse towards which remotion itself is outside itself.”[16] Here too, like Gadamer and Husserl, Heidegger uses the term horizon not to refer to the limits of understanding, but to the conditions of understanding that always draw us toward new understandings.

What then should we make of the earlier quotation the early quotation where Gadamer seems to present horizons as limits? Recall there he writes that

Every finite present has its limitations. We define the concept of “situation” by saying that it represents a standpoint that limits the possibility of vision. Hence essential to the concept of a situation is the concept of a “horizon.” The horizon is the range of vision that includes everything that can be seen from a particular standpoint. [17]

A standpoint limits what we can see, but it is not the horizon that is presented as the limit. Rather the horizon is “everything that can be seen.”  In the same paragraph he says that “horizon” refers to the way our “range of vision is gradually expanded” and that to have a horizon “means not being limited by what is nearby, but to see beyond it.”[18] After looking at Husserl and Heidegger’s usages of the term we can see better what Gadamer has in mind, not limits, but what is graspable within particular limits.

Notice the key difference between Husserl and Gadamer’s use of the term. Perceptual horizons are always present; horizons for understanding may not be and may require work to be acquired. When we are trying to understand another person or a text, we need to have some idea of the horizon in which the subject matter is intelligible to the author or speaker. This does not mean that we need to know the intentions of the speaker or the author (though knowing them might help) as horizons are objective features of the conceptual environment that make a subject matter intelligible, not subjective features of the mind of the speakers.

When we try to understand a text, we do not try to transpose ourselves into the author's mind, but, if one wants to use this terminology, we try to transpose ourselves into the perspective within which he has formed his views.  But this simply means that we try to understand how what he is saying could be right.[19]

To understand the objectivity of meaning horizons, consider the perceptual analogy again: the fact a house is perceived as having a back side is not a fact of the intentions of the viewer, but a fact of the nature of houses. Likewise, the fact a sentence is intelligible is not a fact of the intentions of the speaker, but of the nature of language and of the subject matter of the sentence. Having background knowledge of the subject matter parallels Husserl's internal horizon and having background knowledge of the context of the sentence parallels Husserl's external horizon. The key to a proper interpretation lies in acquiring the proper horizons.

The two horizons, the inner horizon of the subject matter of the sentence and the outer horizon of the historical context of the sentence, are not unrelated. When Descartes says that the body is like a machine, we need to not only know what kind of thing a body is, and what kind of thing a machine is, but what kind of thing they would have thought a body and machine were in the seventeenth century in order to see what this sentence means. If we find that in Descartes's context, calling something a machine was a sign of praise for its complexity and craftsmanship, then we get an idea of what the claim “human bodies are machines” means, which would be very different from the one we'd get if we knew that calling something a machine in Descartes' context meant it was sure to break at some point. Also if we find that in the seventeenth century “body” referred to all physical objects, then again we have a quite different understanding from the one we'd get if “body” referred only to animal or human matter. So the outer horizons itself inform us of the appropriate inner horizons, of the possible subject matter of the sentence. When dealing with texts, the outer horizon is what Gadamer refers to as the historical horizon. The candidates for the meaning of a sentence are given by the historical context.

The inner horizon informs us of the outer horizon as well. All the examples above are plausible ways to think of the body (as something that's complicated, as something that breaks, as being fundamentally related to all other physical things, or as fundamentally different from inanimate things); we'd reject an interpretation of “machine” if it led us to nonsensical conclusions about the body. For example, if we were to find a seventeenth century association between the word “machine” and the phrase “the use of iron in construction” we'd rightly conclude that is not the sense of “machine” in Descartes’s claim that the body is a machine. The relevant interpretive contexts, the historical horizons, are those that produce possibly true sentences. Notice that in practice determining the meaning of a sentence and determining how a sentence might be true are not two separate operations. This insight is necessary for understanding how the fusion of horizons is properly called a fusion.

If we take Gadamer's understanding of horizons as the beliefs that make possible the understanding of a sentence, then we can see what he means by the fusion of horizons. Horizons fuse when an individual realizes how the context of the subject matter can be weighted differently to lead to a different interpretation than the one initially arrived at. Either new information, or a new sense of the relative significance of available information leads, at the very least, to an understanding of the contingency of the initial interpretation, quite possibly to a new understanding of the subject matter, and ideally to a new agreement between the two parties about the subject matter. In any case, the original understanding is surpassed and integrated into a broader, more informed understanding. One's horizons are broadened; we have a new perspective on our old views, and maybe new views as well. This is the meaning of the “fusion of horizons.”

It should be clear, then, how Hirsch is mistaken. He has taken horizons to be fixed limits on understanding, and although it is true that at any moment the horizon is a limit on what can be understood, it doesn’t require extraordinary effort for the horizon to change and new understandings arise. The problem with thinking of horizons as limits is that one thinks of surpassing limits as crossing the horizons. In fact, as we've seen, to know what lies beyond a horizons doesn’t require crossing the horizon, it simply requires moving toward the horizon, or more aptly, moving to higher ground so the previous horizon is included in a broader horizon. Our intellectual horizons change whenever we learn something new or when we weigh differently what we already know, and these changes do not require crossing beyond the limits of our understanding. Gadamer says, “the historical movement of human life consists in the fact that it is never absolutely bound to any one standpoint, and hence can never have a truly closed horizon. The horizon is, rather, something into which we move and that moves with us. Horizons change for someone who is moving.”[20] To conceptually “move” requires simply gaining new information or new insights, a common enough event.

Moreover, because the horizon is our sense of perspective, Gadamer confesses that in fact there is only one conceptual horizon, the world, and everything we encounter is encountered within that very abstract horizon. This is analogous to Husserl's one, unifying spatio-temporal perceptual world. As all statements are intelligible as statements, even if they are intelligible as belonging to a foreign language, so “everything contained in historical consciousness is in fact embraced by a single historical horizon.”[21] Gadamer naturally asks the question this raises: “If there is no such thing as distinct horizons, why do we speak of the fusion of horizons and not simply of the formation of one horizon?”[22] His answer is that speaking of one horizon, although true, underplays the tension that arises between divergent background beliefs that shape interpretations of the subject matter and that form the conditions for disagreement. Emphasizing horizons in the plural shows how these divergent beliefs operate to organize our experiences and interpretations, and consequently how a change in these beliefs amounts in a change in the organization of out experiences and interpretations.

I began by showing the need to arrive at an account of dialogue that doesn’t require agreement to be successful, but still presents a version of agreement for the right reasons as ideal. Gadamer's account of the fusion of horizons, properly understood, hits this middle ground. It is not necessarily agreement about the subject matter, but it is, by Gadamer’s definition, a shared understanding about the subject matter.  While the participants in a dialogue still aim at agreement, the dialogue is a success if each has acquired a new perspective on which to see the subject matter—a perspective informed by that of the interlocutor and grounded in new insights about the subject matter in the context of the claim. That is, the participants have acquired a new understanding of how they originally valued the evidence such they arrived at conflicting conclusions about a subject matter. They have either revised their evaluations, or they have set their evaluations in larger context of possible evaluations. Either way they have acquired a new understanding of the subject matter and of the contingency of their own perspective on it.



                                                                                                            David Vessey

                                                                                                            University of Chicago

[1]  In “Gadamer's Theory of Interpretation,” an appendix to his Validity in Interpretation (New Haven: Yale University Press, 1967).

[2] Validity in Interpretation, p. 254, italics his.

[3] Truth and Method (New York: Crossroad, 1989), p. 302, italics his. Hereafter TM.

[4]  Dordrecht: Kluwer Academic Publishers, 1990.

[5] Ideas I, §27, §44. Italics his.

[6] Den Hague: Martinus Nijhoff, 1973, p. 44.

[7] Phenomenology of Perception (New York: Humanities Press, 1962), p. 68.

[8] TM, 245.

[9] “What is Truth?” in Hermeneutics and Truth, edited by Brice Wachterhauser (Evanston, Ill.: Northwestern University Press, 1994), p. 42. This work from 1957 contains his first published use of the phrase the fusion of horizons.

[10]  “What is Truth?” p. 44.

[11] TM, 305.

[12] TM, 302.

[13]  For example see “The Hermeneutics of Suspicion,” Man and World 17(3/4) (1984), p. 316.

[14] Being and Time, translated by Joan Stambuagh (Albany, NY: SUNY Press, 1996), p. xix. For a detailed discussion of the development of the concept see Theodore Kisiel’s The Genesis of Heidegger’s Being and Time (Berkeley: University of California Press, 1995), pp. 446–451,

[15] Being and Time, p. 334.

[16] Bloomington, IN: Indiana University Press, 1982, p. 267, italics his. This text contains both Heidegger’s most extensive use of the term horizon and his last use of it as a technical term.

[17] TM, p. 302

[18] TM, p. 302.

[19] TM, 292.

[20] TM, 304.

[21] TM, 304.

[22] TM, 306.

Back to David Vessey's Homepage