A picture held us captive. And we could not get outside it, for it lay in our language and language seemed to repeat it to us inexorably.
–Philosophical Investigations 115, 48e
The greatest obstacle in our consideration of alternate philosophical standpoints is often the philosophical standpoint we happen to occupy. We are never fully in command of our assumptions, never fully able to deny the momentum of philosophical habit, and as a result, our standpoint becomes the transparent lense through which we evaluate any alternative. We end up criticizing alternate ways of philosophizing according to the implicit criteria and the default assumptions belonging to our own way.
When rationalists, for instance, accuse contextualists of ‘performative contradiction’ this is precisely what happens. Against the contextualist claim that ‘everything is contingent,’ rationalists typically argue that this claim is either objective and thus incoherent, or subjective and so unworthy of serious consideration. For rationalists, this is obviously the case, and they are bewildered when contextualists blithely continue to advance this claim. But this criticism is only ‘obvious’ from the standpoint of rationalism. For the contextualist, the claim that everything is contingent is a particularly powerful move in a game, not a self-sufficient statement that either is or is not truthfully related to the world. The rationalist’s criticism of this claim, in other words, follows from the very assumptions and criteria that the contextualist repudiates. It begs the question.
Rather than demonstrate the out and out incommensurability of standpoints, however, such examples merely show the difficulty involved in evaluating standpoints other than our own. And like many difficulties, it can be mitigated by shifting our tactics. We can, for instance, evaluate the adequacy of an alternative according to intuitions that seem independent of our standpoint. Or we can evaluate an alternative diagnostically, either point out its problematic presuppositions or show how it might be construed as a more limited version of our own standpoint. Or we can evaluate an alternative internally, and demonstrate either the apparent absurdities it engenders or its inability to meet its own criteria.
None of these tactics, however, guarantee that we have escaped the default assumptions belonging to our own standpoint. So long as we presuppose far more than we can suppose, nothing can prevent our critique from inadvertently begging the question. Given this, perhaps the best way to initially approach alternatives is to not be critical at all, to suspend, as best we can, our commitment to our own standpoint. Only a thorough and charitable understanding of an alternative’s assumptions can prevent us from tacitly substituting our own. We must, as the old adage goes, first walk a mile in an alternative’s shoes. The point is, after all, to find the best outlook possible, and if there is even a possibility that our way of philosophizing (and even those who deny having a philosophical position, like the post-structuralists or quietists, nonetheless have a distinctive way of ‘philosophizing’) precludes our ability to genuinely evaluate alternatives, then it seems irrational to do otherwise.
What follows is an invitation to walk a brief mile a new pair of shoes–to first understand, and then critique a philosophical approach which might be called ‘comparativism.’ Throughout, my aim has been to offer the reader not only a sense of the overall position, but of its apparent ability to dissolve certain long-standing philosophical problems and to diagnose the errors of its competitors. What follows is only an outline, however. Most of the arguments and explanations are abbreviated, either in the interests of economizing the presentation, or because I have yet to work out the broader implications of this approach. *
Our knowledge of the world can never be understood unless we understand how our cognitive relation to the world is mediated. Since our knowledge of the world is irrevocably bound up in language and perception, this amounts to saying knowledge can never be understood unless we understand how our relation to the world is discursively and perceptually mediated. Until recently, the primary candidate for this mediation had been representation. What better way to explain mediation than through the postulation of intermediaries? Language and perception mediate our relation to the world, the assumption was, by furnishing us with representations of it.
But as the ‘re’ in representation might lead us to suspect, this simply repeated the problem of cognitive mediation at two different levels. A representation is simply a presentation that repeats a presentation. The question, accordingly, became one of what mediates our relation to this repeated presentation and of what mediates the relation of this repeated presentation to the ‘original’ presentation of the world. Representation compounded rather than answered the mystery of how our relation to the world is mediated.
The representationalists, however, remained undaunted. As a relation between mental entities, and apparently unburdened by the ‘problem of illusion,’ our relation to representations was considered to be unmediated. The real pressing problem, it was thought, was that of the relation between representations and the world. A handful of daring souls embraced the reductio and suggested there was no relation because there was no independent world, while others attempted to explain the relation in terms of ‘correspondence,’ as though the relations of resemblance between representations must somehow be analogous to the relations of resemblance between representations and an unwitnessed world.
Then Wittgenstein asked a simple question: How do linguistic representations constrain our applications of them? If words are ‘pictures,’ in other words, then how do these pictures bear on the normative dimension of language use? Through his notorious ‘gerrymandering argument,’* Wittgenstein showed that representations cannot constrain applications: the same representations can have different applications depending on their interpretation. Adding ‘rules’ to representations, he argued, is futile, since different applications can be made to accord to the same rule depending on how the rule is interpreted. Adding an interpretation to a rule and a representation is likewise hopeless, since this interpretation can itself be interpreted in different ways. And so on.
In effect, Wittgenstein’s question returned our scrutiny to the mystery of our relation to representations. Whatever our relation to representations might be, Wittgenstein contends, it possesses no bearing on our correct or incorrect application of those representations. From the standpoint of the representationalist, this is nothing short of catastrophic. If, despite our allegedly direct relation to representations, those representations in no way constrain our applications of them, then representations apparently have nothing to do with those applications that qualify as knowledge. Representations, it would seem, do not in fact mediate our discursive relation to the world.
In the years following Wittgenstein’s ‘normative turn,’ normative context has gradually become the primary candidate for the mediation of our discursive relation to the world. For Wittgenstein and his contextualist successors, what constitutes knowing the world is simply what counts as ‘knowing the world’ within the context of concrete linguistic practices. The question of the truth and error of our claims, that is, question of which claims constitute knowledge, is a normative one. Knowledge of the world resides in those cognitive linguistic performances sanctioned as such by a given network of linguistic practices.
Like representationalism, the relation between us and this new medium is thought to be immediate. Rather than governing our discursive practices from above, norms are thought to be constitutive of our practices. But also like representationalism, the question of how our relation to this medium translates into a relation to world proves problematic. Although normative contexts certainly belong to the world, it is difficult to understand how they afford us objective knowledge of the world. From the standpoint of our discursive language games, for instance, we say we know more about the world than the ancient Sumerians not because our linguistic performances stand in a more privileged relation to the world than ancient Sumerian linguistic performances, but because they stand, not surprisingly, in a more privileged relation to our normative context.
Where representationalism had saddled us with a ‘veil of representations,’ one might say contextualism now afflicts us with a ‘veil of performances.’ In contextualism, our practical relation to one another exhausts all other relations, including our cognitive relation to the world. Knowledge of the world becomes a species of performance within a normative context, a ‘move’ in a ‘game.’ By providing the setting of the moves we make in the games we play, the world does in fact play an integral role in our determination of cognitive value. We frequently use how things are in the world independent of our language games, our nonlinguistic context, to determine the propriety of our cognitive moves. The role played by the world, in other words, is a practical one. The world mediates our attribution of cognitive value to our linguistic performances, just as the world mediated ancient Sumerian attributions of cognitive value to their linguistic performances. The problem, however, is that the propriety of that use or mediation is likewise determined by our normative context. For instance, if confronted by our competing truth claims, the ancient Sumerian could complain that we are using our nonlinguistic contexts to attribute cognitive value in an inappropriate way. On the contextualist account, participants in a discursive practice merely instrumentalize the world according to the demands of that discursive practice. The world as it is apart from this intrumentalization recedes behind the performative veil, and with it, the possibility of objective knowledge vanishes. Our nonlinguistic contexts are transformed into a parade of performance-serving features–a kind of ‘arena world.’ The game not only devours all, it devours everything.
In contextualism, the mediation of our cognitive relation to the world by normative contexts leads to the instrumentalization of the world in the service of our claims. As with representationalism, the mediation of our cognitive relation to the world renders the world as it is independent of us inscrutable. We are entombed within the ‘game’ as thoroughly as we were entombed within the picture–perhaps even moreso, given that the mysteriousness of objective knowledge in representationalism becomes outright spuriousness in contextualism. An apparent advantage enjoyed by the model of mediation offered by contextualism, however, lies in the fact that our cognitive linguistic performances are situated in the world rather than in our heads. In contextualism, the possibility of objective knowledge of the world evaporates in the midst of the objective world, and with it our entitlement to such platitudes as ‘We know more than the ancient Sumerians,’ or, even more troubling, ‘The holocaust was wrong.’ Is it possible to amend the game in such a way that this ‘in situ objectivity’ can be exploited? Or does the account of the objectivity of our claims require an entirely different model of the way our cognitive relation to the world is mediated? Many have already abandoned the picture, do we need to move beyond the game as well?
In his landmark work, Making It Explicit, Robert Brandom attempts, among other things, to capitalize on this in situ objectivity (and so ‘pierce the veil’) given his detailed analysis of the ‘social ontology’ of what Sellars had called the ‘game of giving and asking for reasons.’ Unlike Wittgenstein, Brandom is explicitly concerned with the normative significance of cognitive linguistic performances for individual performers within concrete perspectivally structured contexts. In Wittgenstein our perspectives on performances and circumstances within the confines of a given language game are largely taken for granted. For Brandom, however, our perspectives constitute the fundamental structural feature of the game. Without players, after all, there is no game.
According to Brandom, the game consists of a perspectivally structured series of normative ‘undertakings’ and ‘attributions’ within a largely implicit, but sometimes explicit, network of inferential relationships. Within the game, our linguistic performances find their significance in the attribution of ‘deontic statuses’ from the perspectives of other players. Claims made are understood as commitments undertaken against a horizon of further, inferentially articulated commitments. Our entitlement to these commitments, that is, their cognitive adequacy, depends on the normative status attributed to them by other players from the standpoint of their own repertoire of discursive commitments. Moves made in the cognitive language game, in other words, are understood in terms of the perspectivally attenuated expectations and estimations of inferential cognitive adequacy–what Brandom calls ‘deontic scorekeeping.’
Brandom is aware, however, that an account of the game in terms of the ‘I-Thou’ relation between perspectives seems to aggravate rather than relieve the problem of objectivity. Every player is at once a referee in Brandom’s game, and no player enjoys the status of an ‘absolute scorekeeper.’ This means the normative status of our claims, our entitlement to those commitments undertaken, is always a contingent matter of negotiation. Rather than explaining, as any account of ‘knowledge talk’ must, how discursive practices have mediated the explosive growth of knowledge since the Enlightenment, Brandom’s perspectival account seems to render it even more mysterious, given that our entitlement to our commitments seems to deteriorate into a matter of ‘his word against mine.’
Brandom responds to this problem by deflating objectivity, redefining it in effect, into something that pertains to the form of the game as a ‘feature of discursive intersubjectivity.’ In fact, he locates the possibility of objectivity in the very feature of the game that seems to render objectivity problematic: its perspectival structure. The perspectival difference between commitments undertaken as true and commitments attributed as taken to be true, Brandom claims, allows for the “permanent possibility of a distinction between how things are and how they are taken to be” (597).
Brandom formalizes the possible ways in which objectivity might be short-circuited into a series of conditionals. If his social ontology can defeat these conditionals, he believes, then he will have proved that objectivity in his account of the game does not deteriorate into a ‘mere privileging’ of communal or individual perspective. For instance, the no first-person ignorance conditional, ‘If p then I claim that p,’ is defeated in his account because although we might be committed to both ‘p’ and ‘I claim that p,’ it can be the case that we are not entitled to ‘p’ and yet remain entitled to ‘I claim that p.’ What this means, Brandom argues, is that claims about the world cannot be reduced to claims about commitments. How things are, in other words, cannot be reduced to how things are taken to be. The game thus allows for possible “genuine fact-stating discourse” which involves “practices of assessing claims and inferences according to their objective correctness–a kind of correctness that answers to how things actually are, rather than how they are taken to be, by anyone” (607).
But the same perspectival structure, Brandom fails to note, also allows for the permanent possibility of a collapse of this distinction. An irony of Brandom’s demonstration, for instance, is that it actually relies on the collapse of how things are into how things are taken to be in order to fund their distinction. To say we are no longer entitled to ‘p’ is to say we realize ‘p’ is not in fact how things are, but merely how things were taken to be. Our entitlement to the consequent, ‘I claimed that p,’ remains unaffected, because although no longer entitled to p as an act of referring to the world, we remain entitled to the act of referring to that reference. The difference in entitlements that defeats the conditional and so prevents the collapse of how things are into how things are taken to be arises directly from the collapse of that same distinction.
The differences in entitlement that assures the nonequivalence of how things are and how things are merely taken to be is a difference in entitlement from the standpoint of the player, and so does not amount to the ‘permanent possibility’ of objectivity tout court, as Brandom supposes, but simply the permanent possibility of objectivity from the standpoint of the player. All Brandom shows is that, from the standpoint of players in the game (since this the standpoint from which the conditionals are defeated), claims about how things are cannot be globally reduced into claims about how things are taken to be. But disproving the necessary equivalence of how things are and how things are taken to be from the standpoint of players does not disprove their possible equivalence, and thus does nothing to dispel the problem of objectivity, the problem that our accounts of how things are may turn out to be accounts of how things are merely taken to be. It does not disprove either the fact that what counts as ‘how things are’ at one point in the game will often count as ‘how things were taken to be’ at another, or that from a standpoint exterior to the game, every instance of ‘how things are’ might be a mere instance of ‘how things are taken to be.’ All he demonstrates, in effect, is that from the standpoint of the player the possibility of objectivity cannot be ruled out, which stands to reason, since the arbitration of objective correctness is presumably what the game is all about.
It should come as no surprise that these infallibility and omniscience conditionals are defeated by instances ‘being incorrect,’ or that Brandom’s normative model of discourse, which depends upon the possibility of ‘being incorrect,’ defeats them. But the question at stake is not the trivial one of whether the structure of the game allows communities and players to be counted incorrect, but one of whether the structure of the game allows communities and players to overcome the problem of objectivity. As Brandom acknowledges, the objective correctness of our claims is a correctness that answers to how things are, that is, a correctness that does not merely answer to ourselves or to other players. An objectively correct claim, in other words, would be a claim we are entitled to by virtue of how things are, a claim sanctioned by the world rather than by others. In order to overcome the problem of objectivity, then, Brandom’s game must explain how the correctness or incorrectness of our claims is determined by the world independently of the scorekeepers.
But Brandom’s account, like Wittgenstein’s, is a ‘norms all the way down account.’ On such accounts, a player’s locutions are correct only insofar as they are counted as correct by other ‘scorekeeping’ players, while the world is simply the arena of our cognitive ‘plays,’ something we either use or ignore in order to ‘score’ with our locutions. The correctness of locutions about how things are, in other words, depends entirely on whether scorekeepers concur with our use of ‘how things are.’ How things are independent of this use, that is, independent of the game, has absolutely no bearing on the correctness of our claims about how things are. One might say the scientist, for instance, is constrained by the world in the same way a carpenter is constrained by his tools. For the carpenter, the propriety of his use of tools is entirely determined by the network of practices we call carpentry. Likewise, the propriety of the scientist’s use of ‘empirical observations’ is entirely determined by the network of practices we call science. Our perspectives on the world can be used in innumerable different ways, the respective proprieties of which are entirely determined by the scorekeepers. Rather than using the game to win the world, as Brandom seems to suggest, on his account we can only use the world to win the game. [footnote the ‘mere power struggle passage’]
The ‘objective correctness’ of our claims answers, not to the world as Brandom suggests, but to the normative status scorekeepers attribute to our instrumentalizations of nonlinguistic contexts. Different scorekeepers in different practices mean different assessments of objective correctness, whether they have access to the same arsenal of observations or no. The permanent possibility of ‘objectivity’ that arises from the perspectival structure of Brandom’s game, therefore, must be merely subreptive, a way perhaps to force players to conform in their instrumentalizations and so prevent the game from fracturing into an explicit competition of interests.
Objectivity eludes Brandom because, like Wittgenstein, he remains trapped behind the veil of performances.
If all significance, including the significance of the world, reduces to use, then the significance of the world reduces to its instrumentalization by our practices, to the role it performs in our games, the propriety of which is determined by our scorekeepers. By construing the game in explicitly perspectival terms, however, Brandom inadvertently foregrounds the debilitating irony that dogs all ‘game accounts’ of discursive practice. Since a perspective on the linguistic performances of others is a necessary condition of the game, it simply follows that our perspectives are perspectives on how things are independent of us. And yet, despite this perspective on an objective world, objectivity nevertheless evaporates in game accounts of linguistic practice.
The inability to translate perspectives on an objective world into discursive objectivity suggests that something is drastically wrong with the game account of the way perspectives, world, and language constitute discursive practices. The primary symptom of this problem is found in the inability of game accounts to make sense of determinations of comparative cognitive adequacy. If our entitlement to the claim that we know more than the ancient Sumerians is secured by how things are with our scorekeepers rather than by how things are with the world, then we are, in fact, never entitled to say we know more than the ancient Sumerians. We can only say that what counts as ‘cognitive’ for contemporary scorekeeping contexts differs from what counted as ‘cognitive’ for ancient Sumerian scorekeeping contexts. When we ask whether we know more than the ancient Sumerians, we are asking whether the claims counted ‘objective’ by our present scorekeeping context are ‘more objective’ than the claims counted as ‘objective’ by other scorekeeping contexts. Since our contemporary scorekeeping context entirely determines what counts as ‘objective,’ claims generated by our scorekeeping context will always be counted as ‘more objective’ than those generated by others. From the standpoint of their scorekeeping context, the Sumerians will be just as entitled to claim they ‘know more’ than us.* Determinations of comparative cognitive adequacy, then, simply identify players as belonging to one scorekeeping context rather than another, and say nothing, in fact, about whether we actually know more about the world.
In fact, on the game account, we can never affirm any platitude of the form ‘x knows more now than y knows at time b.’ We cannot, for instance, say we know more now than we did yesterday, since what entitles us to such a claim is our present scorekeeping context, that is, a different scorekeeping context.1 Certainly, we assume that our present scorekeeping context is the same as the scorekeeping contexts of moments ago, but we are entitled to this assumption only by virtue of our present scorekeeping context. The game not only devours all and everything present, it devours all and everything past as well. Other players, the world, and even the past only possess ‘inferential valence’ by being incorporated into our present practices, that is, by being used in some way. As a result, comparisons of the cognitive adequacy of locutions secured by different scorekeeping contexts always amounts to a reduction of those contexts to our present scorekeeping context. We know more about the world, not because of our cognitive relation to world as mediated by our discursive practices is in fact more comprehensive, but because, as products of our present scorekeeping contexts, our present claims are better adapted to our present scorekeeping contexts. When our cognitive relation to the world is mediated by the game, in other words, ‘knowing more’ about the world has nothing to do with the world.
Determinations of comparative cognitive adequacy require a tertium quid, something common in reference to which we can say we ‘know more,’ in order to be intelligible. When we say we know more than the ancient Sumerians, we are saying we know more about something common to our respective discursive practices. This ‘something common,’ presumably, is the world. This is why the game account’s inability to make sense of determinations of comparative cognitive adequacy is symptomatic of the game account’s inability to translate perspectives on an objective world into discursive objectivity. The ‘something common’ that provides a necessary condition for determinations of comparative cognitive adequacy is actually a given in the game account. Despite the differences in scorekeeping contexts, the game affords both us and the Sumerians a shared perspective on a common objective world. Nevertheless, the game cannot account for determinations of comparative cognitive adequacy, because present scorekeeping contexts monopolize the determination of objective correctness by determining the propriety of our instrumentalizations of the world. Our perspectives on a common world deteriorate into perspectives on different sets of tools.
Determinations of comparative cognitive adequacy require that our respective perspectives on a common world somehow constrain the correctness of our claims independently of our respective scorekeeping contexts. Put differently, determinations of comparative cognitive adequacy require what we expect from ‘discursive objectivity’ in general: that it allow for, in Brandom’s words, “a kind of correctness that answers to how things actually are, rather than how they are taken to be, by anyone.” In this sense, the ability to account for determinations of comparative cognitive adequacy constitutes an important ‘objectivity condition’ for general accounts of the way our cognitive relation to world is mediated. Since the game, even in its exhaustive Brandomian incarnation, fails this condition, some alternative is needed.*
In the aftermath of the normative turn, however, the problem has been to find an alternative to the game that does not run afoul some variant of Wittgenstein’s critique of representationalism. Any account of the world/word/us relation must explain the essentially normative character of language use, and as it stands, the game appears to be the only model able to do so. But in satisfying Wittgenstein’s ‘normativity condition,’ the game transforms the world into an arena, and so fails the ‘objectivity condition’: it cannot make sense of determinations of comparative cognitive adequacy. What we need, therefore, is an alternative account that satisfies Wittgenstein’s normativity condition without running afoul of the objectivity condition. What we need, in other words, is an account of the way our cognitive relation to the world is mediated that somehow completely accords with the game (insofar as the game meets the normativity condition) without being exhausted by the game (insofar as the game fails the objectivity condition).*
At least so far, these normativity and objectivity conditions have amounted to the Scylla and Charybdis of theories of discursivity. Accounts that meet the normativity condition tend to run afoul of the objectivity condition while accounts that meet the objectivity condition tend to run afoul of the normativity condition. We began with the need to understand how our cognitive relation to the world is mediated. Given the failure of the two major analogical models of this mediation, the picture and the game, to simultaneously meet these two conditions (and after decades of refinement and elaboration, no less), the time has come to explore an entirely different analogical model of this mediation. *
As mentioned, Brandom’s primary problem lies in his inability to translate our perspectives on an objective world into discursive objectivity. Since this problem resides in the way our perspectives on the world are ‘taken up’ by the game (as instrumental moments of itself), the question of how perspectives are taken up by discursive practices provides a promising starting point for an alternative model of cognitive mediation. Any answer to this question, however, requires a prior understanding of what it means to possess a perspective on the world in the first place. Our alternative model, in other words, must involve an ontological account of perspectives.
As the structural cornerstone of the game, one might expect that varying ontological accounts of perspectives will commit us to varying ontological accounts of the game, and that, conversely, these social ontologies will commit us to certain ontologies of perspective. For his part, Brandom makes at least three cardinal ontological assumptions regarding perspectives: 1) he assumes we possess a perspective on the world as it is independent of us; 2) he assumes an intraperspectival structure that exchanges the Cartesian picture of perspectives on concepts for the Kantian picture of perspectives grasped by concepts (636); and 3) he assumes the classic distinction between what is merely perspectival (how things are taken to be) and what is the case (how things are).
The difficulties involved with assuming (1) are notorious, and need not be rehearsed here. The difficulty Brandom faces by assuming (2), however, is somewhat more subtle. Brandom insists, quite properly, that understanding concepts as implicit material inferential roles avoids the Kantian dualism of concept and intuition, form and content. The dissolution of this dualism, however, seems to be incompatible with his commitment to the Kantian picture of concepts grasping perspectives. Since concepts simply are contents, and since concepts grasp perspectives and not vice versa, this also means that contents grasp perspectives. But if conceptual contents grasp perspectives, how is it that perspectives grasp the world by grasping conceptual contents? Who grasps what? And for that matter, what does ‘grasping’ mean?
But if objectivity is only problematically given by (1), and if (2) muddles the relation between perspective, world, and language use, it is (3), perhaps the least contentious of these ontological assumptions, that precipitates the transformation of the world into an arena. Like locutions, our instrumentalizations of our perspective on the world in the service of our locutions are ‘objectively correct’ only from the perspective of scorekeepers within some scorekeeping practice; they are ‘objectively correct,’ that is, only by being taken as such by scorekeepers. *
In (1), the assumption that our perspective is a perspective on an objective world, we secure the possibility of access to something independent of our ‘always already’ contextualized perspectives. An ontology of perspectives must exploit the possibility that a perspective is always a perspective on the world.
With (2), the question of the relation of perspectives to concepts, contents, and the world, we are reminded that the fundamental goal of any philosophy of language is not simply to clarify the ‘word-world relation,’ but to understand the relation of word, world, and us. An ontology of perspectives must explain how linguistic performances mediate the cognitive relation of our perspectives to the world.
And since (3), the distinction between what is merely perspectival and what is the case, seems to be the fulcrum of Brandom’s difficulty, we need to consider a revision of this distinction. An ontology of perspectives somehow must either rework, dissolve, or overcome the dualism of what is and what is taken to be, without surrendering the objectivity promised by (1).
A NEW ONTOLOGY OF PERSPECTIVES
In answer to (1), I will argue that perspectives are always from somewhere on something transcendent, that this ‘something transcendent’ is the world, and that perspectives are incoherent otherwise. Perspectives are constitutively intentional, which is to say, object directed. This is why a perspective is always a perspective ‘on something.’ But perspectives are also constitutively privative. A perspective on something is always a limited perspective. To say we possess a perspective ‘on something’ is to say we possess only one ‘angle’ among many on the same something. Our perspective on a table in a room, for instance, is only coherent as a perspective in contradistinction to all other possible perspectives on the table. This is simply what it means for a perspective to be ‘from somewhere’ and ‘on something transcendent.’ Implicitly, we understand that our position in the room limits our view of the table, and that the table, accordingly, transcends our perspective on it. Our perspective on the world, in other words, is a perspective on a transcendent world because the position of our perspective in the world limits us to one of many different perspectives on the same world. We are framed by the very world we frame in our perspectives.
Our perspectival relation to the world, in other words, is mediated by our position in the world. Since experience is the fulcrum of knowledge, and since our experience of the world consists in the transformations of our perspectives on the world, one might say that in some sense our cognitive relation to the world is mediated, not by representations, nor by our scorekeeping contexts, but by our perspectival position in the world. How much we know of a table in a room, whether or not it possesses three or four legs, for instance, often depends upon our position in the room. The more positions we occupy, we like to say, the more we know the table.
The fundamental idea here is to use the positional mediation of our perspectival relation to the world as an analogical model for all senses of our cognitive relation to the world, including the linguistic. A perspective’s ‘from somewhere,’ on this model, need not mean ‘from some spatio-temporal position.’ The ‘position’ of a perspective on the world may be specified in any number of ways: historically, culturally, physiologically, linguistically, and so on. A perspective’s cognitive relation to the world, on this model, is mediated by the ‘sum’ of these positions. They specify the ‘from somewhere’ of our perspectives, and thus the manner in which our cognitive relation to the world is mediated.
Consequently, all cognitive differences in perspective must be understood as differences of position, not as differences in representation. Our perceptual modalities, for instance, demarcate a certain position in the field of all possible perceptual modalities–a position in ‘perceptual space.’ Colour is not a ‘mental construct,’ a representational by-product of our parochial physiology, but rather a modality of our perspectival relation to the world. The different qualitative features of our perceptual perspective are best thought of as different angles on the world from different ‘perceptual positions.’ Likewise, the different cultural features of our perspective are best construed as different angles on the world from different ‘cultural positions’ in ‘cultural space.’
Whenever we speak, as we often do, of ‘cultural perspectives’ (‘the traditional point-of-view’), ‘discursive perspectives’ (‘theoretical standpoints’), in short, whenever we relativize claims to some context, we rely on our implicit understanding of the positional mediation of our cognitive relation to the world. We implicitly understand that contexts mediate our cognitive relation to the world, but we have hitherto lacked the ontological resources needed to make this understanding explicit. This model of ‘positional mediation’ doubles as an account of the relation between context and perspectives. Contexts, on this account, are understood as perspectival frames. Contexts are how things are pertinent to a perspective on how things are. They attenuate perspectives by specifying the position that ‘frames,’ that is, mediates, their relation to the world.
This constitutes a drastic shift from the two central analogical models of the tradition: the ‘picture’ and the ‘game.’ Where the former casts us adrift by mediating our relation to the world with reified intermediaries, the latter casts us adrift by mediating our relation to the world through arbitrary normative contexts. In this account, however, our relation to the world is mediated by our position in it. Since our perspective is always a perspective on an objective world, no matter what our position, we are never in danger of ‘losing the world’ behind a veil of perceptual representations or linguistic performances. The only way to lose the world, on this account, is to leave it.
In answer to (2), I will argue that linguistic performances frame concrete perspectives on the world.
Using Brandom’s ‘tactile trope,’ one might say linguistic performances ‘grasp’ us in such a way that we ‘grasp’ the world. Although Brandom maintains the identity of concept and content, ‘being grasped’ and ‘grasping’ would seem to be completely different things. Since we are presumably ‘what is grasped’ by concepts, and since the world is presumably ‘what is grasped’ by us, we need a model where the grasp language games have on us simply is the grasp we have on the world.
But what does it mean to be ‘grasped’ by a linguistic performance, and how does this ‘being grasped’ allow us to transform ‘our grasp of how things are’?
The idea here is to consider perspective in first-person terms, as a ‘field of experience,’ rather than in third-person terms, as an individual standing in some relation to how things are. The problem with this latter standpoint is that it relies on what might be called ‘thing-thing relationality.’ The relation between perspective and world is external, a relation between two independent things. The first-person standpoint, on the other hand, is characterized by what might be called ‘figure-field relationality.’ Our perspective constitutes the field, or frame, within which certain features of the world are ‘figured.’ In spatial terms, for instance, the way a perspective frames any given feature of the world is determined by its position. Our view of the world is always a concrete view ‘from somewhere.’ Occupying different positions in a room allows us to frame the room’s various features differently. As the frame of our perspective is transformed, so is our grasp of ‘how things are.’
This, I will argue, provides an apt analogy for understanding what it means ‘to be grasped’ by linguistic performances. Linguistic performances, as mentioned, frame concrete perspectives on the world. To make a move in a language game, in this sense, is to literally move one’s interlocutors to a certain position in what might be called ‘semantic space,’ to transform their perspective in a manner analogous to redirecting them to a different position in a room.
The ‘content’ of any given linguistic performance, accordingly, is nothing other than the concrete perspective it frames. We are literally the content of what we hear. This model forces us to exchange the venerable ontological distinction between ‘subject’ and ‘object,’ which characterizes our relation to the world as an external relation between things, for a revised, loosely Heideggerean, understanding of ‘being in the world.’ Rather than identifying ourselves with any ‘thing,’ such as the ‘mind,’ ‘consciousness,’ and the like, we must identify ourselves with our relation to the world. In first-person terms, we simply are our positionally mediated perspective on the world. Being, as Heidegger says, is always concrete ‘being in the world.’ Our contextually framed relation to the world is simply who we are.
This model, not surprisingly, also forces us to abandon the contextualist characterization of meaning as ‘use.’ When we use language, we frame our own perspective and the perspectives of others in different positions vis a vis the world. The meaning of any given locution is simply the matter of fact, contextually (that is, positionally) mediated perspective it frames. Aside from this perspective, locutions are mere noise. The normative dimension of language use, accordingly, must be understood in terms of positions rather than performances. The correctness of linguistic performances, on this model, depends upon whether it frames our relation to the world in an appropriate way. Normative contexts sanction positions, not performances. ‘Moves in a language game’ are at once moves in the world.
Since the meaning of any statement is simply the momentary perspective on the world it frames, any linguistic transformation of a perspective’s position in the world is at once a transformation of a perspective’s view on the world; the perspective/world relation is internal. Thus, to be ‘grasped’ by a linguistic performance is at once to ‘grasp’ the world. To say a linguistic performance frames a perspective, is to say linguistic performances frame a certain relation to the world. ‘Scheme’ and ‘content,’ on this account, are simply different sides of the same perspectival coin.
In answer to (3), I will argue that there is no absolute distinction between ‘how things are’ and ‘how things are taken to be.’ To say a perspective is always a perspective on the world is to say a perspective is always a perspective on ‘how things are.’ Since knowledge is the default on this account, the problem, accordingly, becomes one of explaining ignorance. If the analogical model of the positional mediation of our cognitive relation to the world makes the objectivity of that relation the default, the problem becomes one of accounting for the cognitive differences between perspectives.
Since our cognitive relation to the world is mediated by our position in it, it stands to reason that the ‘truth’ or ‘falsity’ of our perspectives, linguistically framed or otherwise, becomes a matter of occupying different positions. Since positions are relative, however, questions of the truth and falsity of various perspectives would seem to be likewise relative. But these positions are not ‘points’ in some abstract space, instead they are concrete perspectival positions on a concrete world. Some positions in a room offer more comprehensive vantages on the room than others, just as the position of being sighted offers a more comprehensive perspective on the world than being blind.
The distinction between how things are and how things are taken to be, on this account, is merely the comparative distinction between more and less comprehensive perspectives on different features of the world. A perspective can always see more of how things are–more than another and more than itself. Even though a perspective is always a perspective from somewhere, and therefore a limited perspective, perspectives are transformed over time. The ‘from somewhere’ of our perspective is framed by all the other ‘from somewheres’ we have previously occupied. This is why, for instance, our adult perspectives on the world are more comprehensive than our childhood perspectives.
Previous perspectives on the world frame subsequent perspectives. When we move from position to position in our room, for instance, we see that the table in fact possess four legs instead of three, and we realize that our previous perspective was idiosyncratic. Since perspectives can only be understood as privative in contradistinction to other perspectives, the possibility of more comprehensive perspectives on the world is an ontological feature of perspectives.
This is significant in at least two respects. First, it suggests a crucial way in which ‘knowing more’ (possessing a more comprehensive perspective) is independent of communal assessments normative adequacy. The relation a perspective possesses to the world is a matter of fact. Although normative contexts determine which perspectives we ought to occupy, they do not, in the end, exhaustively determine the actual comprehensiveness of those perspectives. Even children raised by wolves may ‘broaden their perspective’ on how things are. Since our survival in the world depends upon the comprehensiveness of our perspective, and since the ‘survival instinct,’ the imperative to survive, is prior to our participation in normative practices, it is always possible to make a correct move that is at once individual and unprecedented. In other words, more can be better independent of our practices, and because of this, it is always possible to ‘throw off the yoke’ of our practices and transform them in light of how things are.
Second, it suggests not only how it is we come to ‘know more,’ but also the powerful way in which language facilitates our knowing more. Although always ‘context bound,’ perspectives are also always ‘context transcendent.’ Even though at any given moment we possess only a single angle on a table in a room, that angle is framed by both previous angles and the possibility of subsequent angles. The actual and possible transcending of contexts, in other words, is always part of our context. And the transcending of context always amounts to knowing something more, no matter how trivial that ‘something’ might be.
Language, on this account, is our principle instrument of context transcendence. Linguistic performances, the suggestion is, frame concrete perspectives on different features of the world. By framing perspectives, linguistic performances reenact actual angles on the world. Since this world transcends our perspectives, the relation reenacted is in certain respects the same relation. An elk’s cry of alarm affords the herd a relation to a wolf in the absence of any perceptual perspective on it. This is the foundation of all linguistic communication: the reenactment of certain perspectives on the world through linguistic performances. Language, in other words, enables inter-perspectival context transcendence.
In addition, language differentiates, organizes, and condenses our perspectives on the world. Our perspectives are framed by our previous perspectives. The comprehensiveness of any given perspective is determined by the ‘from somewhere’ of our perspective, by its position. This position, in turn, is partially specified by the way previous perspectives implicitly frame it. Language, on this account, dramatically expands the possibilities of this framing of perspectives by previous perspectives, and thus drastically enhances our capacity to occupy more comprehensive positions, that is, to transcend contexts.
The association of perspectes on the world with concrete linguistic performances in the world allows for 1) the condensation of perspectives into an inventory of repeatable types possessing various syntactically and materially constrained combinatorial permutations; 2) the instrumentalization of perspectives; and 3) the semantic mobility of perspectives. When using language, we exploit the combinatorial possibilities of linguistically condensed perspectives to enact or reenact perspectives apart from whatever nonlinguistic position we occupy. Even though perceptually and practically restricted to the ‘here and now,’ our perspective can be ‘from anywhere’ linguistically. Not only can we actively pursue perspectival comprehensiveness through language, we are able to transcend our position in time and space in order to do so. The same linguistic features that make inter-perspectival context transcendence possible enable intra-perspectival context transcendence as well.
MEETING THE CONDITIONS
Given this rough ontology of perspectives, we are now in a position to understand, in outline, how the analogical model it provides for the mediation of our cognitive relation to the world might simultaneously satisfy our objectivity and normativity conditions. With reference to the former, the model clearly allows for determinations of comparative cognitive adequacy. The question of whether we ‘know more’ than the ancient Sumerians, becomes the question of whether we possess a more comprehensive perspective. Since ‘comprehension’ is an ontological feature of our perspectives, the comprehensiveness of our perspectives is a matter of fact, just as the question of whether one has possessed and integrated more perspectives on a table is a matter of fact. Comprehension, on this account, is recognized, not merely ‘attributed.’
We secure this recognition from others by reenacting this comprehension for others, that is, by reenacting actual cognitive relations to the world. Since this reenactment consists in the instrumentalizations of our perspectives on the world, however, and since the proprieties of these instrumentalizations are determined by our normative context, this recognition would seem to count for little more than attribution, insofar as we only recognize what our parochial scorekeeping context deems ‘recognizable’ as ‘comprehensive.’ It could be, for instance, that the ancient Sumerians actually possessed a more comprehensive perspective, and that we are simply unable, due to our scorekeeping contexts, to recognize it as such.
Everything seems to collapse back into performances. Scorekeeping contexts, however, also designate a position in the world on the world, and as such, can be comparatively parochial or comprehensive, depending on the actual comprehensiveness of the perspectives they legislate. Scorekeeping contexts must answer to the ‘presocial’ imperative to survive, and since our survival depends upon the comprehensiveness of our perspective on how things are, scorekeeping contexts, whether they belong to us or to the ancient Sumerians, must also answer to how things are. The propriety of our scorekeeping contexts, then, is determined by the propriety of the ‘comprehensive views’ they legislate, which is in turn determined by the survival imperative: we must know the world in order to successfully navigate it. We instrumentalize our perspectives on the world, in other words, not to merely secure our place in the game, but to secure our place in the world.
Not only, then, do we know more than the ancient Sumerians in fact, we can, on the basis of a shared world and a shared survival imperative, show them that we know more, and thus force them either to abandon their scorekeeping contexts where they challenge the proprieties of our instrumentalizations, our to amend them. This is why one modern doctor in ancient Sumer could transform their entire culture, while an ancient Sumerian doctor in our day would likely be banned from practicing medicine.*
The question of whether this model abides by Wittgenstein’s normativity condition, however, is more complicated, and will occupy much of what follows. For the moment, it is important to keep in mind that perspectives must be understood in their first-person sense. We ‘look’ through rather than at perspectives. When we frame perspectives, that is, take a perspective on our perspectives in order to ‘look at’ them, they become inert, something externally related to questions of correct and incorrect application, just as when we frame rules, make them explicit, they become external to our rule-governed practices. The question then becomes one of how linguistically framed perspectives, as something distinct from their application, determine the correctness of their application. The same regress that paralyzes ‘legalistic’ accounts of rules, would then also paralyze our understanding of the relation between perspective and normativity.
But just as there is a way of ‘grasping’ a rule that does not involve after the fact framing, there is a way of grasping content that also does not involve after the fact framing, but which is exhibited in what we call ‘having a perspective on the world’ in actual cases. Since we are our perspective, there is a sense in which our perspective constitutively exceeds our ‘grasp.’ Certainly we can take a perspective on our perspectives, ‘look at them’ if you will, but the perspective we must look through in order to look at our perspective always exceeds us. The position that implicitly frames, in other words, always outruns the position explicitly framed. A good analogy for this is found in our inability to ‘look at’ the edge of our visual field. Although our visual field frames a limited view of the world before us, the frame itself is occluded.
When perspectives are understood in first-person terms, that is, in terms of their ‘occluded frame structure,’ the question of how isolated, linguistically framed perspectives ‘determine’ the propriety of their applications becomes incoherent. First, because there is no such thing as an ‘isolated perspective.’ The occluded frame that specifies our matter of fact relation to the world is constituted, in part, by the sum of our retained and projected perspectives. And second, because normativity is, as already suggested with reference to the normativity condition, nothing apart from our perspective (which is not say that it is any less real). The occluded frame that specifies our relation to the world is constituted, in part, by our normative context, which is to say, by our ‘normative position.’ Perspectives are always already normative perspectives.
The Wittgensteinian formulization of the ‘normativity condition,’ in other words, simply does not apply to a perspectival position account of cognitive mediation. This, however, in no way relieves the burden of explaining the normativity of language use. The story to be told here will be in part a genetic one. The ‘occluded frame structure’ of perspectives, it turns out, allows us to understand what normativity is, why it is our performances are ‘correct’ or ‘incorrect,’ rather than say, ‘healthy’ or ‘unhealthy.’
COMPARATIVISM: AN OVERVIEW, AND SOME FURTHER COMMITMENTS
The dividends of this ontology of perspectives are at once mundane and dramatic. On the one hand, it simply enables us to affirm two common-sense platitudes: that we possess perspectives on the world and not just on representations of it, and that the world constrains our perspectives on it independently of our normative context. On the other hand, however, it explains how finite and contingent creatures such as ourselves could possibly possess, extend, and exchange objective knowledge of the world. It allows us to overcome, in other words, the vexing opposition between truth and context.
Truth and context are reconciled by making explicit the implicit epistemological structure of perspectives. Perspectives are ontologically epistemological. When we use, for instance, the metaphors of a ‘view from nowhere’ and a ‘view from somewhere’ to describe the distinction between ‘objective’ and ‘subjective’ cognition, we are relying on our implicit understanding of perspectives. On this account, the relation between the ‘view from nowhere’ to objective cognition and the ‘view from somewhere’ to subjective cognition is literal rather than metaphoric. Since all our perspectives are perspectives on the world, all our perspectives are ‘objective’ in some sense. And since all our perspectives constitute one angle on the world among others, all our perspectives are ‘subjective’ in some sense. Cognition is at once objective and subjective. On this account, ‘subjective’ and ‘objective’ are comparative terms, merely two sides of the same perspectival coin. Cognition is ‘objective’ to the extent our perspectives transcend our ‘view from somewhere,’ and ‘subjective’ to the extent they do not. Cognition is not a hit or miss affair. In the same way our position in a room conditions our knowledge of its contents, our position in the world, our context, conditions the comparative truth of our claims. Claims are more or less true, not either true or not.
This ontology of perspectives, in other words, suggests a comparative theory of truth. Truth, on this account, pertains to the comparative context transcendence of the perspectives framed by our claims–the degree to which they approach ‘the view from nowhere.’ The more comprehensive the perspective framed by our claims, the more true those claims are. Truth is the measure of the actual context transcendence of perspectives, whether this transcendence is manifested in the inferential ‘coherence’ of a group of linguistically framed perspectives on the world, the ability to make generalizations about the veracity of statements made by others, the ability of hypotheses to bear up under experimentation, or the ability of maxims to allow the satisfaction of our expectations in action. Our ontology of perspectives affords us a glimpse of the ‘elephant’ behind all these intuitive aspects of truth: context transcendence.
An important question raised by comparative truth, however, is one of how it bears on its bivalent discursive deployment. Bivalence, I will argue, is an artifact of the ‘occluded frame structure’ of perspectives, the sense in which we stand both‘somewhere’ and ‘nowhere’ at once, and the practical need to ‘be decisive.’ By transforming the muddy matter of arbitrating between different perspectival positions, all of which are ‘true’ insofar as they constitute more or less comprehensive perspectives on the world, into a matter of arbitrating true and false perspectives, bivalence allows us to be decisive, that is affirm and deny perspectives outright without worrying about ‘degrees of affirmation.’ Bivalent truth, in other words, is a subreptive convenience. We must be decisive in order to survive.
The all important question, however, is one of how we arbitrate the ‘comparative context transcendence,’ or truthfulness, of our linguistically framed perspectives. The answer, quite simply, is that we continue chasing our perspectives come what may, continue experimenting, acting, reflecting, and arguing, and that this is always a ‘messy and retail business.’ Only as our perspective broadens over the course of time can we definitively determine whether a certain perspective on the world is ‘more true’ than another, whether we ‘know more’ about cancer, say, than we did ten years ago.
Knowledge is the default in this ontology of perspectives. So long as we possess a perspective on the world we know something of the world. The interesting question, on this account, is not one of whether we know anything at all, but rather how it is we come to know more. When all positions are positions of understanding, the question becomes one of how and why some positions provide a more thorough understanding of the world than others. The central philosophical question pertaining to this ontology is at once a practical question.
*Since our assurance of the objectivity of our knowledge lies in our knowing more rather than simply knowing differently, the position arising from this ontology of perspectives might be called ‘comparativism.’
Comparativism is an anti-representational realism that partially revises but largely adheres to the central insights behind Wittgenstein’s normative turn. Comparativism qualifies as anti-representational through the contention that the representational (as opposed to the positional) mediation of a perspective’s linguistic or perceptual relation to the world is incoherent. In representationalism, our perspective retreats from the world and becomes a perspective on representations of the world. Not only can it be shown that this model renders perspectives incoherent, it can be shown how it is we come to think in ‘representational terms’ in the first place.
Comparativism qualifies as a realism–a version of direct realism in fact–through the contention that our perspectives are perspectives on an actual independent world. The problem of illusion, upon which so many versions of direct realism founder, can be shown to dissolve in the distinction between parochial and comprehensive perspectives. The skeptic’s challenge to direct realism, on the other hand, can be defused by pointing out the triviality of the skeptic’s claim: the only thing skeptical arguments show us is that our relation to the world is constrained by our perspective. For the skeptic, not seeing all of everything means not seeing anything at all.
Comparativism harmonizes this realism with the normative turn by recharacterizing Wittgenstein’s ‘moves in a game’ as ‘moves in the world.’ Far more than Brandom, Wittgenstein focuses on the social ontology of the game at the expense of the perspectival ontology that makes the game possible. In the absence of performers, the constituents of the game become contextualized performances. The world, which at once provides the background and falls into the foreground of our perspectives, becomes pure background in Wittgenstein, part of the context that allows us determine the propriety of linguistic performances–an ‘arena world.’ In comparativism, however, it is not just the propriety of linguistic performances that is adjudicated, but rather the propriety of what those performances do, that is, the way they frame our relation to the world.
We condition statements by saying they belong to such and such a ‘standpoint’ all the time. We implicitly understand, in other words, that perspectives play an intimate role in cognition. But without an explicit understanding of the ontology of perspectives, we rely on default assumptions, implicit perspectives on perspectives, that may or may not lend themselves to the integration of a more comprehensive perspective–to understanding.
From the standpoint of this ontology of perspectives, which I have dubbed ‘comparativism,’ comprehensiveness provides the central criterion of adequacy. The theoretical strength of comparativism depends upon its ability not only to explain various facets of cognition and language use, but also to diagnose, ‘encompass’ if you will, competing philosophical standpoints. Comparativism, I believe, offers not only a more comprehensive picture of ‘how things are’ with cognition and language, but also explains why it is we so readily fall into the traps of representationalism and contextualism.
Since it rarely serves to hazard a ‘gestalt shift’ in philosophy outright, I will gradually work my way toward a definitive statement of the position, confining myself for the moment to a single question: What is the relation of truth to context?
This relation, I think, is encapsulated in the metaphor of a ‘view from nowhere.’ To apprehend the world ‘from nowhere’ is to have an unconditioned view of the world, to apprehend the world as it is independent of any contingent context. To apprehend the world ‘from somewhere,’ on the other hand, is to be embedded in contexts, and although this does not necessarily mean that our view is a mere ‘construct,’ it does mean that any justifications we might adduce for the truth of our view will depend upon our context, upon our role as participants in some social practice.
Taking the metaphor of a view from nowhere as a clue, then, the relation between truth and context might be construed as perspectival. To apprehend the world from somewhere is to have a restricted rather than an unrestricted view of the world. What this means is that context does not necessarily preclude the possibility of comparative truth. So far as we possess a perspective on the world, and not on some representational construct, one might say that we always possess a view that more or less approximates the truth. The more unrestricted our perspective on the world is, that is, the more it approximates the view from nowhere, the more true it is. And yet, since our perspectives are always bound to the ‘here and now,’ it would seem that any individual perspective, no matter what the context, is always as restricted as any other perspective–and the notion that all perspectives are equal, as Plato pointed out long, long ago, renders the very notion of truth unintelligible. If truth is to be explained in terms of perspective, then, we need to account for cognitive differences between perspectives, the way views from somewhere can approximate the view from nowhere, despite our restriction to the here and now.
And yet, even if it were possible for a view from somewhere to approximate the view from nowhere, the decisive problem would be one of justification. Since the justification of any particular perspective we might have on the world depends upon the social practices that constitute our context, the issue of arbitrating the comparative truth of our perspectives would seem to depend upon the shape of whatever ‘language-game’ we happen to be playing. If such justification is to be more than simply relative to our language-games, not only would we need to account for the way perspectives on the world ‘hook up’ with our language, we would also need to account for the way the content of those perspectives constitutes a source of normativity independent of our games.
We use terms such as ‘frame of reference,’ ‘standpoint,’ ‘position,’ ‘point-of-view,’ ‘vantage,’ and so on all the time in philosophical discourse, and typically without any explicit understanding of what they mean. In the most general terms, what follows is an attempt to use what has hitherto been regarded by many as the enemy, ‘mere perspective,’ and show that it provides a way to preserve many traditional intuitions regarding truth without relinquishing more recent, and quite powerful, insights regarding context once it and its relation to language are properly understood.