(Ph.D., University of Nottingham) is Assistant Professor of Theology at Emmanuel University, Oradea, Romania, E-mail:
ABSTRACT: There are several tensions present in George Lindbeck’s postliberal theology. One of these is between realist intuitions and a non-epistemic ac- count of truth, on the one hand, and a social- constructivist non-realism with regard to theological statements. Theology is relegated to the status of second-order discourse, while first order language comprises the practices, rituals, vocabulary of a reli- gion. I am challenging this intermediary status of religion with the help of Donald Davidson’s critique of the dualism between scheme and content. Since there isn’t much to be made of the notion of epistemic intermediaries (either as ‘content’ or as ‘scheme’) it does not make sense to continue to speak either in terms of realism or of non-realism. This clears the way for treating theology as cognitive.
I wish to challenge what has become a widely ac- cepted distinction in contemporary postliberal theol- ogy. The distinction between first order truth claims and second order grammatical reflection stems from the application of the linguistic metaphor to religion and from allowing theological methodology to have what Rowan Williams has called a “territorial cast of imagery.”1 Describing religions as ‘conceptual frame- works’ may seem to have given due weight to holis- tic considerations. It may appear that careful consid- eration has been given not only to the propositional aspect of a religious system, but also and more im- portantly to the expressive, the symbolic and to the performative side as well. Traditions, rites, liturgies, worship have gradually emerged as forming not sim- ply the outer garment of an essentially propositional
Distinction: A Critique of the
Dualism Between Orders of
George Lindbeck, Donald Davidson, Bruce D. Marshall, holism, philosophy of language, theological method, postfoundationalism, justification, ontology.
system, but as central to its very identity. Such a re- newed interest in form, rather than simply content, in style, rather than simply substance, gave the lin- guistic metaphor its naturalness, which is my task here to demystify. The emergence of this central metaphor followed hot on the heels of transforma- tions within philosophy as well, where empiricism and idealism were retreating before the onslaught of the philosophy of language, both British/ Austrian and American.
Hastily assuming the felicity of the analogy, phi- losophers and students of religion abandoned the former theoretical orthodoxy and engaged in the
‘thick description’ of religions. Like languages, these were taken to be structural wholes, possessing a vo- cabulary, a grammar and internal grammatical rules for these. They are transcendental schemes which condition perception, which legislate what may count as a good argument, what is most important for truth, indeed they decide what is most important in life. Building on a Kantian picture of perception2, it was supposed that the religious scheme provided the necessary mechanism by which raw experience, of the sacred as well as of the mundane, was
organised. Together with the notion that there may be other rival conceptual frameworks dividing reality in their own incommensurable ways,3 the linguistic turn seemed to drift inevitably towards relativism.
Yet there is something wrong with this picture. I want to argue that postliberalism, as the theological orientation which builds most forcefully on the dual- ism of first and second orders of language, has be- trayed some of its best holistic possibilities in its be- ing seduced by a certain unconvincing view of
language. The understanding of language as a reified whole, as a unitary structure which is constructively active at all times, or as a scheme applicable to any content, must give way to a more flexible, more au- thentically holist view. This clarification of holism is mostly rooted in the philosophy of Donald
Davidson, but it is congenial to the work of the later Wittgenstein, of Michel de Certeau and of Pierre Bourdieu. These thinkers, in their own personal way, have reoriented the attention of philosophy to the possibility of transgressing tradition, boundaries, language, fixed regimes of control, without however reinventing foundationalism or any kind of naïve re- alism. I shall restrict my comments to how Donald Davidson can assist postliberals in clarifying some in- ternal inconsistencies.
2. The two orders of language in postliberal theology
There is in postliberalism an unapologetic ten- dency to assume the primacy of its own textual tradi- tion. In part this mentality stems from a dissatisfac- tion with the ‘cognitive-propositionalist’ neglect of the element of social construction present in knowl- edge. For postliberals such as George Lindbeck, knowledge and meaningful speech are only possible from within a conceptual scheme, or respectively language: “There are numberless thoughts we can- not think, sentiments we cannot have, and realities we cannot perceive unless we learn to use the ap- propriate symbol systems.”4 Meaning and reference
are distributed by the conceptual scheme itself. The assumption which must be noted from the outset is that the meaning of these conceptual schemes is available quite independently of the actions of hu- man agents which inhabit this textual world. Al- though Lindbeck professes a dialectical relationship between scheme and experience,5 it is quite clear that the relationship is dominated by the conceptual scheme, as he himself admits.6
Within such religious schemes, two types of truth claims are possible. On the one hand there is what Peter Ochs calls an ordinary use of language, a com- mon sense referral to the world. This is accom- plished by way of first-order statements such as
‘Christ is Lord’. The truth of this statement is onto- logical, if and only if it is correlated with a form of life ontologically adequate to ultimate reality. The first order claims of a religion are those which make up the ‘vocabulary’7. These are the statements of lit- urgy, prayers, stories in the sacred texts, acts of wor- ship etc. All these constitute the performative aspect, which may assume specifically cultural forms, but which is regulated by the normative context of Scrip- ture and tradition.
Seldom performative, the second order level comprises the reflective claims of theology, which does not make any ontological truth claims, but merely reflects and corrects the performance by drawing attention to its regulative inner grammar.
The function of theology is, in Kathryn Tanner’s words, to step in when practice falters8. It is in this sense a meta-linguistic clarification of practice. There are two different moments of meta-linguistic clarifi- cation, which, argues Ochs, derive from Lindbeck’s
rule theory of doctrine. Ochs reads the rule theory as “a theory about Scripture’s translational tenden- cies.”9 On a first, intra-textual level, there is an ex- plicit distinction between ordinary linguistic practice (first order) and meta-linguistic practice, through which we organise, correct and adjust ordinary prac- tice. On a second, inter-textual level, an implicit dis- tinction is made between natural language in gen- eral and what Ochs calls the “transformational language of the Gospel text, or Scripture”.10 Accord- ing to this second principle, Scripture itself func- tions within a natural language as a meta-linguistic correction, a translation of the mundane world into a sacred world.
Ontologically true claims can only be made on the level of ordinary use of language. There is a bi- conditional explication of this truth: it depends on an intra-systematic coherence with the form of life (its grammar, given by its doctrines, its practices) and, secondly, it depends on the ontological ad- equacy of the scheme itself. This is Lindbeck’s re- placement concept for that of propositional truth.
Religious systems are neither true, nor false, but they simply make true or false statements possible, like languages.11 The contextualist assumption is clear: a statement must cohere with a form of life, in order to be meaningful. In other words, communi- cation is possible only where language is present.
And quite clearly, Lindbeck’s idea of language is dominated by the notion of structure.
However, Lindbeck goes one step further than the contextualist in suggesting that meaning is con- ditional not only upon an existing framework, but also upon an adequate framework. The categories of
the given language must be adequate to expressing features of reality. Echoing positivist notions of meaningfulness, Lindbeck concludes that should the language prove inadequate, first order claims are not simply false, but meaningless.12 The dubious deci- sion to call such talk meaningless reveals a certain half-heartedness about Lindbeck’s attempt at non- foundationalist theology. It seems that, as Sue Patterson convincingly argues, he has great difficul- ties in completely relinquishing older realist notions of what makes a statement true.13 If one of the great- est faults of realism, according to its critics, is that of making truth radically non-epistemic14, than
Lindbeck’s notion of adequacy involves precisely that kind of non-epistemic situation. The adequacy of a religion is something which one can only guess at by checking its “assimilative power”15. Yet it re- mains decisively non-epistemic in that it will only be revealed in the eschaton. This notion of Lindbeck will have the sure effect of appeasing conservative critics to some extent, especially since it is accompa- nied by a notion of the absoluteness of a “final reli- gion”.16 Yet it should serve as a warning that there are important contradictions in Lindbeck’s system.
The question of the sort of reference theory which Lindbeck may have in mind must be raised at this point. This partially non-epistemic nature of sig- nification suggests that reference is an all or nothing affair. Either the religious conceptual scheme and the form of life associated with it is adequate to ulti- mate reality, in which case reference is possible by right performance and internal coherence, or the framework itself may be inadequate, thus rendering meaningless all religious speech acts. Besides the
semi-Pelagian implication of such a view,17 it presup- poses a descriptive (or intentionalist) theory of refer- ence, whose shortfalls it inherits. The descriptive theory of reference (Frege and Russell) holds that the reference of a term is whatever object in the world fits the description abbreviated by that term.
Saul Kripke has pointed out the main difficulty with such a theory, namely that the more false beliefs we have about something, the less in touch with the world may be.18 Lindbeck would fall under Rorty’s critique of theories of reference as something we ought to have intuitions about. For him reference, as an all or nothing affair, describes a mysterious rela- tionship which obtains between parts of language and parts of reality. Should our language prove ad- equate, our terms may have reference. But if it isn’t, then we have all been babbling about nothing.
This theory of reference also entails another theoretical position, namely, that different tradi- tions, with different beliefs and customs, may be re- ferring to different things. Or, if we are to take Lindbeck’s suggestion about the superiority of one religion, only one tradition refers, while all the oth- ers don’t. Yet given Lindbeck’s ecumenical focus, it is precisely this possibility which he wants to avoid.
The trick, in my interpretation, is to attach reference not so much to the outer, cultural and performative aspect of a religion, but to its inner grammar and most especially its Scriptural narratives. Reference is ruled by such a grammar and by the way in which the Gospel narratives themselves identify Jesus and God. This in turn requires that despite all
performative diversity there is an unchanging core to the Christian religion and that core is its grammar.
Such a notion of reference presupposes what Davidson names as the third dogma of empiricism, the dualism of scheme and content19. It is the scheme, our linguistic contribution to knowledge, which dictates reference. The scheme is highly organised, as Lindbeck himself testifies: “religions are seen as comprehensive interpretive schemes, usually embodied in myths or narratives and heavily ritualized, which structure human experience and understanding of self and world.”20 The linguistic na- ture of such a scheme gives the impression that its meaning is fixed, that its grammar is clear and stable.
The community of believers ensures that the mean- ing of the scheme is quite well kept21. Indeed, it can be acquired and it need be acquired prior to any truth claiming, prior to any discussion of legitima- tion. One becomes religious by learning the lan- guage of that religion, by becoming skilled in the practices, customs, worship of a specific religious community. Once one has internalised this frame- work, only then can one make truth claims: “one must be, so to speak, inside the relevant context;
and in the case of religion, this means one must have some skill in how to use its language and prac- tice its way of life before the propositional meaning of its affirmations becomes determinate enough to be rejected.”22 It appears, then, that the dependence of truth claims on a highly structural language is matched by the belief that such a language can be learned quite apart from one’s engagement with the world or indeed with ‘ultimate reality’.
Once schemes are taken to condition reality in this fashion, it is only a small step to the supposition that we live in different worlds. Thus ‘God’ becomes
a concept which we have relaxed into accepting23, differing from one community to another, one that we simply learn to accept. Theology delivers no new knowledge to the believer, but reflects on past per- formances.
3. Reconnecting meaning and truth holistically
A pervasive disjunction between meaning and truth functions in narrative theology. Frei’s
prioritising the meaning of the Gospel narratives ahead of their truth and Lindbeck’s separation be- tween lexical meaning and second-order theological truth point to the relative semantic stability of the re- ligious framework. This emphasis on meaning fol- lows in the wake of the advent of linguistic philoso- phy. Far from taking us much beyond modern epistemology, however, most philosophy of lan- guage amounts to a mere linguistic translation of Kant.24 Theoretical interests are now oriented to- wards discovering the human contribution to knowl- edge in the form of the meanings of sentences. The basic philosophical intuition at work here is that dis- covering how language works means in fact discov- ering the limits of knowledge. It is not difficult to trace the path of what we may call a ‘receding real- ism’ of a constant narrowing of the ‘objective’ in epistemology. This degenerative myopia ‘started’ (in as much as a causal chain can be said to have a proper beginning) with the rejection of objective knowledge of divine reality and fell back upon the
certainty of the senses or of the eternal substances.
Then these were abandoned too, while all epistemic forces converged on ‘mind’ or rather ‘spontaneity’.
Then came the realisation that mind itself is a cul- tural and temporal product. All that we have now is language. Yet a distinctive common characteristic of all these phases is the belief in the availability of the epistemic locus, be it God, sensorial experience, mind, mental objects and now language.25
‘Language’ now intervenes between agent and world as the new Given. In theology this distinction takes the form of separating inquiry into the struc- ture of religion from investigating the truth of its statements. This distinction is fiercely guarded since it is only ‘insiders’, those who have first learned the language, understand the spirit and follow to the let- ter, that should attempt to probe into its truth. This obviously involves the gamble that the scheme and the language we strive to learn is inadequate after all. I have remarked the oddity of this non-epistemic aspect within Lindbeck’s project. But the notion that knowledge is relative to a scheme has a grasp on us only if we think that there is something like ‘mean- ing’ which one must first grasp before one gets to the world.
This relativising of truth to meaning can only make sense if the distinction between questions of meaning and those of fact is valid, and this is a very difficult position to hold in a post Quinean world.
The notion that language expresses something in- ner, which is captured by definitions and by distin- guishing between the essential and accidental prop- erties of an object is no longer tenable after Quine’s attack on the two dogmas of empiricism. Thus it no
longer seems plausible to distinguish between “what one is talking about” and “what one is saying about it”26, or between statements which are true in virtue of meaning and those which are true in virtue of em- pirical considerations.
The adoption of holism entails taking leave of such dualisms. No longer expressing something in- ner, meaning becomes a property of behaviour, as Dewey put it, and thus it can no longer serve as foundation of philosophical inquiry. ‘Meaning’ be- comes a congratulatory term which points to the success of communication. In the Anglo-saxon philo- sophical climate which is now so indebted to Quine, meaning explains relations between sentences, rather than relations between sentences and mental objects. For both Quine and Davidson a theory of meaning must be thoroughly empirical theory.
There is no reason to suspect that Lindbeck would disagree with this. Insofar as he doesn’t he remains a holist. But what he does not realise is the conse- quence of making meaning an empirical theory.
Since the only way to write a theory of meaning is by avoiding a circular account which presupposes se- mantic content and by finding the way in which for a particular language sentences are used, to under- stand meanings, intentions, propositions, linguistic content generally requires treating them as involved with our moving around in the world. If meaning is a property of behaviour, such a ‘meaning’ cannot be individuated except by paying attention to the ac- tions it is involved with.
We can only summarise Davidson’s argument here. Since behaviour is all we have to go on, we must treat people as being generally right. We thus
stabilise truth, while making meaning problematic.27 A theory of meaning cannot simply be content with giving descriptions of the form s means that p for a language L. The connector “means that” is ambigu- ous. It is this mistaken assumption about the stabil- ity of meaning which encouraged philosophers to base reference on it. This is also the reason why Quine’s project of radical translation is not good enough for Davidson.28 Translation merely provides for moving from one known language into another known language by giving a sentence-for-sentence translation, but it says nothing about how to gener- ally interpret people’s utterances. We must,
Davidson and Rorty suggest, treat truth as basic and resist giving explanations of truth on the basis of meaning. The way to do this is to explain ‘means that’ with the help of the concept of truth. To summarise: a theory of meaning should not seek to relate sentences to mental entities, but provide a translation of sentences into the theory-language or meta-language. Yet such a translation should not employ the connector ‘means that’, which is itself the object of theory. Thus s is true if and only if p.
The sentence ‘is true if and only if’ gives a sufficient account of what is involved in meaning. It involves speakers assenting to something in the world. Mean- ing and communication is only possible by treating people as reacting in the same way to the same ob- jects we are reacting to in the world.
3.1. Scheme-content dualism and justification
Davidson’s re-connection of meaning and truth, his firm rejection of the separation between ques- tions of meaning from questions of fact, results in his rejection of what he calls the “third dogma of empiricism”. I want to argue that Lindbeck himself inherits the dogma.29 It is the supposition that there may be incommensurable schemes organising reality which gets the scheme-content distinction going.
Davidson’s argument against conceptual relativity is well known and I shall not waste space on it. It boils down to the fact that since untranslatability does not make any sense, the very idea of a scheme
organising reality, derived from the idea of a lan- guage which is true but untranslatable, is pointless.
In fact the very notion that a scheme ‘organises’,
‘fits’, ‘is adequate to’ reality adds nothing of interest to the simple concept of being true.30 Thus a scheme should only be adopted if it were true. The very con- cept of adequacy is revealed as a hopelessly ambigu- ous one. Yet, if we cannot make sense of the idea of untranslatable languages and if translation involves truth, then all we have to go on are our beliefs about what makes a scheme/ theory true or false, rather than adequate, or fitting.
It is not only the ‘scheme’ side which comes in for attack. To speak of sensory experience, rather than simply the evidence in the form of beliefs, or just the facts, urges Davidson, adds nothing new by way of evidence for a theory. The notion of content, as well as the notion of scheme, are unfortunate in-
truders between agents and world. To introduce
‘content’ gets us nowhere, as Bruce Marshall points out.31 It is the belief that one’s experiences have such and such content, rather than the content itself which does all the epistemic work.
It is at this point that Davidson and Marshall sig- nificantly depart from Quine. Davidson believes that Quine has not completely relinquished the third dogma, by his persistence in talk about evidence which is not itself a belief.32 Thus, the scheme-con- tent dualism still fulfils its psychological motivation in Quine: it gives us the impression that there is something on which to base our beliefs. Although Quine drops the notion of direct access, typical of foundationalism, “his talk of statements about the external world ‘facing’ the tribunal of sense experi- ence continues to enshrine an epistemic outlook which turns on a confrontation between linguistic or conceptual scheme and given empirical or experien- tial content.”33 This results from the fact that Quine has permitted certain statements within the web to have a qualitative superiority over other statements due to what we may call ‘experiential proximity’, namely their being closer to the web’s edge. The whole web rests on such ‘observation sentences’ and the further we progress towards the middle the more indeterminacy we get. Ironically, Quine can still be called a foundationalist, in this modest, per- haps unwitting sense. Where Davidson corrects Quine is in dropping the notion of a justification which is not itself a belief. Thus, justification is equally distributed across the web and there is no distinction between the web’s sentences in terms of
their justification by experience. Every belief is equally dependent on other beliefs.
What saves Davidson’s project from the utter and complete relativism by which philosophers like John McDowell34 fear it is haunted, is the transcendental requirement that belief must be on the whole veridical.35 It is this which makes communication possible: we must hold for truth while testing for meaning. If, in other words, we cannot make sense of the truth of an alternative scheme, then we have no grounds on which to ascribe meaning. As Terry Godlove puts it: “No translation, no truth; no truth, no meaning; no meaning, no belief. But without any of this, on what grounds are we alleging an alien conceptual scheme?”36
3.2. The reification of language
Only on pain of separating meaning from behaviour, from truth and from translation that philosophical imagination could be captivated by the linguistic metaphor. Now I will spell out the implica- tion of Davidson’s holism for our understanding of what a language is.
Due to the formal nature of Tarski’s theory of truth, which Davidson applies to get at meaning, other considerations are necessary in order to apply it to particular linguistic performance.37 Holism, in- terpreted in this fashion, does not take meaning lightly, despite all the rhetoric about indeterminacy.
It was in fact the previous, pneumatisch in
Wittgenstein’s parlance, theories of meaning which
tended to consider meaning on the whole fixed.
There was thus the danger of treating linguistic abil- ity as an affair between a speaker and a system of signs. The particular dialogic occasion, with its usual transgressions of grammar and invention of vocabu- lary shows that one has only mastered a language once one knows how to move ahead when all rules have been abandoned. For Davidson, these are not simply deviations from the norm, but they are the es- sential feature of communication.38 In light of this it appears that for Davidson communication does not, as he was mistakenly interpreted, start and end in agreement. Difference is present at the outset as in- centive for communication.39
Since meaning is not fixed, the success of com- munication does not depend on the participants ac- cessing a common set of meanings,40 but on the skill of creating the desired effect in others by using sen- tences in specific ways. We reach understanding when we are able to anticipate the effects on us that others intend.41 If meaning is a generic term which describes behaviour,42 then communication is achieved “where causes converge: your utterance means what mine does if belief in its truth is system- atically caused by the same events and objects.”43
Linguistic ability, consequently, cannot be sepa- rated from our inhabiting the world. If meaning is not learned in advance of inter-subjective interac- tion, prior to being in the proximity of the causes of the interlocutor’s beliefs, then communication can- not be based on conventions one has learnt in ad- vance. Such conventions are indeed helpful in com- munication, but they cannot ever form the basis for it.44 Meaning would be insulated from life and lan-
guage banished outside the world. This inter-subjec- tive element is all important for Davidson, marking his most important departure from Quine, whose epistemology he sees as rooted in the perceptive so- lipsism typical of the empiricist subject.45 Rejecting this ‘deification’46 of language, Rorty and Davidson denounce the persistence of certain epistemic anxi- eties in philosophy, and their creation of founda- tions for thought. ‘Experience,’ ‘mind’, ‘language’
have all been candidates to this privileged status at some point. They must all be returned to the world.
Both language and mind, as Davidson replies to an interviewer, are part of a single inter-subjective ma- trix, along with the world.
The point I am making is Wittgensteinian as well as Davidsonian: linguistic ability is not something to be acquired prior to speaking about the world. The meaning of our linguistic acts is connected to our beliefs about the world, about ourselves, and about what others are saying. In view of these consider- ations, language loses its privileged position and it can no longer be treated as a medium.
The connection to my topic can be rehearsed here. Yarbrough comments that “discourse about discourse has tended toward what I will call
‘linguistification’ – toward more and more emphasis upon language as a medium of force between ‘mind’
and ‘reality’.”47 Perhaps without the complicity of a fixed notion of meaning this would not have hap- pened. Scenarios of different worlds are given greater plausibility by this misconception of mean- ing. Davidson points out that “instead of thinking of these [different cultures] as sort of blocks that are fixed one way or another, we might think of them as
just variance which we understand in terms of what we share and see ourselves as sharing.”48
4. A holist correction of cultural- linguistic theology
Sue Patterson’s observation is spot on: “The feel- ing of discordance or incredulity when it is sug- gested that the whole system might function in a way cut adrift from the claims of its own proposi- tions may be taken as a sign that an inconsistency has been created, some component of the logic skewed.”49 I want to make three short observations.
First, the feeling of incredulity derives from
Lindbeck’s abandoning one of the best postliberal intuitions about the literal sense with respect to doc- trines and theology. These, despite the claims of those who use them, do not ‘refer’ to extra-linguistic reality at all. Secondly, the inconsistency Patterson suspects concerns the presence of elements of both theistic realism and of a radical social constructivism.
I want to suggest that Lindbeck’s realist impulse is due to his half-hearted adoption of holism. Once ho- lism has been adopted across the board, there no longer remains room to state an issue between real- ism and idealism, as Davidson and Rorty50 show. The reason for this is that ‘language’ no longer functions as epistemic intermediary which connects with the world either realistically, discovering it, or idealisti- cally by constructing it. Finally, cutting adrift the system from the claims entails a problematic view of reference and of meaning. Instead of holistically al-
lowing ‘reference’ to be the philosophically un-inter- esting notion51 of ‘what one is really talking about’, Lindbeck wants the reference of a statement to func- tion independently of what people may intend with it.
Lindbeck isn’t alone in his half-heartedness about holism. Quine himself hasn’t grasped the qualitative equivalence of the beliefs within a web.
He qualitatively distinguishes between statements in terms of their location within the web. Some state- ments are closer to the edges of the web, to the world, their meaning is less contaminated by other beliefs. Such statements are generally recognisable by the whole community as non-problematic, be- cause of their proximity to the world. Lindbeck also distinguishes between the qualitatively superior first order statements and the second order theological claims. But what makes the distinction is less clear than in Quine. It may be that first order statements are superior in terms of their being ‘performative’.
But theology can be used performatively as well. An- other criterion might be that of ‘non-identical
repetitivity’, by which they non-identically repeat the claims of Scripture or tradition. This again is not a good criterion since unless we want them to be simple tautologies we need to allow a level of nov- elty within such ‘repetitions’. The task then remains that of justifying the novelty which brings us back to where we started. The fact remains that, unless Lindbeck will clearly formulate reasons why some- thing belongs to the first rather than to the second order, the distinction itself is very hard to flesh out.
It is simply not enough to say that theological state- ments belong to the second order while
performative statements belong to the first, for who is to decide what is a theological statement, or what counts as appropriate performance. Yet Lindbeck wants first order claims to show the limits of the Christian universe. This world is ruled Scripturally.
The task, therefore, is highlighting the border be- tween inside and outside.
Quine attempted the same demarcation by be- lieving the web to be anchored at certain crucial points. It is precisely the positing of such ‘resting points’ which gets the scheme-content distinction going. For if there is an accessible and non-epistemic reality then we have a way to answer the sceptic. But Quine and Lindbeck should have realised that there is no way to say what the world looks like except from within a present theory of the world. Quine understood that the objects of a theory cannot be specified other than by giving translations of that theory into another. For some reason, however, this insight is eluded in his ‘foundationalist’ melancholy.
In light of the above we should be sceptical about the amount of epistemic work done by the distinction between first and second orders of lan- guage. Both Lindbeck and Thiemann, whose ac- count of theological justification needs a treatment in its own right, construct an architectonic model of knowledge. At the bottom lie those statements
which are justified by their coherence with the Chris- tian form of life, and by the adequacy of that form of life itself. Such a vocabulary makes certain first order truth claims, which in the performative act of an ad- herent can be repeated. The truth of these state- ments is qualitatively different from the truth of theological statements, which are a reflection upon
such performances. Lindbeck has thus established a level of foundational discourse which Thiemann later exploited to epistemic ends.52
But if it is only beliefs which justify beliefs, there can be no distinction in terms of sources of justifica- tion. The only relevant qualitative distinction be- tween statements is their truth. This is the theologi- cal upshot of Davidson’s claim that the notions of
‘fitting experience’ or ‘organising experience’ are poor substitutes for the concept of truth. In other words, a belief such as “Christ is Lord” should not be considered a central belief because it belongs cen- trally to the vocabulary of religion which is ‘adequate to’ the working out of God’s salvation. If ‘being true’
is the only criteria for acceptance of a belief, then there can be no philosophical reason why some be- liefs are more central than others. In fact the very notion of a vocabulary and grammar which make up that “configuration of language and activity consti- tuting the Christian faith”53 drops from the picture.
The centrality of certain beliefs turns out as being connected not to some special truth that they pos- sess (ontological truth), while other statements are only intra-systematically true, but to the sociological fact that such a belief is considered to be essential to the identity of a given community. I suspect that Marshall’s insistence to speak in terms of epistemic priorities has more to do with sociological consider- ations about what holds a community together than with epistemic reasons, or else it isn’t authentically holist.
The argument above should not be taken to sug- gest that there is no second-order thinking, that there is no meditation upon language itself. I have
attacked the prescriptive distinction between such orders of language, the notion that whether a state- ment falls in one category can be decided ahead of the analysis of communication. The distinction is prescriptive in that Lindbeck restricts theology to grammar. The world is thrown overboard in this over-enthusiastic pragmatism, to echo Rorty.54 Those adherents of pragmatism and even holism, like
Wesley Robbins and Nancy Frankenberry, which sug- gest that since we cannot make sense of representa- tionalism all that remains is to speak about lan- guage55 only reiterate the dualism of language and world. To say that we can only refer to language is to treat reference as that mysterious ontological con- nection between our words and the world, which ei- ther obtains or it doesn’t. Reference, or ‘aboutness’
can only be discussed on the margin of the commu- nicative situation and not prior to it. Whether theo- logical statements are grammatical claims about lan- guage or ontological truth claims about divine reality depends on much more factors than Lindbeck al- lows.
Within postliberal theology it has been Hans Frei who has paid much more attention to both the theo- logical requirements of a religious theory of knowl- edge56 as well as to the communicative complexity of issues. There can be, he writes, both a grammatical function as well as a cognitive function to doctrines.
For example, the doctrine of justification functions both as a rule, as well as a statement about God which holds true regardless of the attitude of the person articulating it.57 Whether a doctrinal state- ment is to be taken as cognitive or as descriptive of the Christian form of life is not to be legislated be-
fore the actual discursive situation. This seems to co- here well with my Davidsonian argument about in- terpretation theory: the way in which we should in- terpret doctrinal statements can only to a certain extent be informed by our prior theory. We need a passing theory58 in order to make sense of actual us- age.
There is no good reason, therefore, to impose any referential restriction on theology, partly be- cause reference is not all that mysterious and tells us nothing about how we could refute the sceptic, or the unbeliever. Theological knowledge is not para- sitic upon a religion which functions like a language, not if by language is understood a conventional sys- tem of meanings. The very notion of such a language is obsolete and needs to make room for language as an ability to move around in the world. A direct con- sequence, which I cannot elaborate on, is that being skilled in the practices of a religion is not a requisite for understanding its concepts and making pertinent statements about it. This is not to say that the skilled practitioner will not have a more genuine under- standing of its concepts. But it does not follow that adherents of ‘rival’ traditions have no way of under- standing one another.
1 Rowan Williams, On Christian Theology (Oxford:
Blackwell, 2000), p. 29.
2 George Lindbeck, The Nature of Doctrine: Religion and Theology in a Postliberal Age (London: SPCK, 1984), p. 33.
3 Lindbeck, Doctrine, p. 49.
4 Lindbeck, Doctrine, p. 34.
5 Lindbeck, Doctrine, p. 33.
6 Lindbeck, Doctrine, p. 34.
7 Lindbeck, Doctrine, p. 33.
8 Kathryn Tanner, God and Creation in Christian Theology: Tyranny or Empowerment? (Oxford:
Blackwell, 1988), pp. 15-16.
9 Peter Ochs, ‘Rabbinic Pragmatism’, in Bruce D.
Marshall, ed., Theology and Dialogue: Essays in Conver- sation with George Lindbeck (Notre Dame: University of Notre Dame Press, 1990), p. 231.
10 Ochs, ‘Rabbinic Pragmatism’, p. 230.
11 Lindbeck, Doctrine, p. 48.
12 Lindbeck, Doctrine, pp. 48ff.
13 Sue Patterson, Realist Christian Theology in a Postmodern Age (Cambridge: Cambridge University Press, 1999), p. 43.
14 Bruce Marshall, Trinity and Truth (Cambridge:
Cambridge University Press, 2000), pp. 223ff.
15 Lindbeck, Doctrine, p. 131.
16 Lindbeck, Doctrine, p. 49.
17 This can be stated as follows: this makes God de- pendant on a previously existing religious conceptual framework, thus denying the freedom of God’s actions, including the freedom of his revelation. See for an excel- lent analysis George Hunsinger, ‘Truth as Self-Involving:
Barth and Lindbeck on the Cognitive and Performative Aspects of Truth in Theological Discourse’, Journal of the American Academy of Religion, 61/1 (1993), pp. 41- 56, esp. pp. 48ff.
18 Richard Rorty, Philosophy and the Mirror of Na- ture (Princeton: Princeton University Press, 1979), p.
19 cf. Donald Davidson, ‘On the Very Idea of a Con- ceptual Scheme’, in Donald Davidson, Inquiries into
Truth and Interpretation (Oxford: Clarendon, 1984), passim.
20 Lindbeck, Doctrine, p. 32.
21 See Lindbeck’s notion of ‘competence’, Doctrine, p. 35.
22 Lindbeck, Doctrine, p. 68.
23 Alan Keightley, Wittgenstein, Grammar and God (London: Epworth, 1976), p. 45.
24 This is what Rorty called “impure philosophy of language”, Philosophy and the Mirror of Nature, p. 257.
25 Richard Rorty, Objectivity, Relativism and Truth (Cambridge: Cambridge University Press, 1991), pp. 54, 58.
26 Rorty, Philosophy, p. 268.
27 cf. Davidson, ‘Radical Interpretation,’ passim in In- quiries.
28 Davidson, ‘Radical Interpretation’, p. 129.
29 This is also Milbank’s conclusion. John Milbank, Theology and Social Theory: Beyond Secular Reason (Oxford: Blackwell, 1996), p. 387.
30 Donald Davidson, ‘On the Very Idea of a Concep- tual Scheme’, p. 16.
31 Bruce Marshall, Trinity and Truth, pp. 80ff.
32 Donald Davidson, ‘A Coherence Theory of Truth and Knowledge’, in E. Sosa and J. Kim, eds., Epistemol- ogy: An Anthology (Oxford: Blackwell, 2000), p. 157.
33 Marshall, pp. 83f.
34 John McDowell, Mind and World (Cambridge, Mass. and London: Harvard University Press, 1996), pp.
143. See his description of Davidson’s project as “unre- strained coherentism” (p. 143).
35 Donald Davidson, ‘Three Varieties of Knowledge’, in Donald Davidson, Subjective, Intersubjective, Objec- tive (Oxford: Blackwell, 2001), pp. 210, 212; ‘The Method of Truth in Metaphysics’, in Inquiries, pp. 199- 201.
36 Terry Godlove, ‘In What Sense are Religions Con- ceptual Frameworks?’, Journal of the American Academy of Religion 52/2, pp. 289-305, pp. 295f.
37 Davidson, ‘Communication and Convention’, in In- quiries, pp. 270, 272, 274.
38 Davidson, ‘A Nice Derangement of Epitaphs’, in E.
LePore, ed., Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson (Oxford: Blackwell, 1986), pp. 441-442.
39 Stephen R. Yarbrough, After Rhetoric: The Study of Discourse Beyond Language and Culture (Carbondale:
Southern Illinois Press, 1999), p. 183.
40 cf. Quine’s notion of the “Myth of the museum”
(Quine, ‘Ontological Relativity’, The Journal of Philoso- phy 65/7 ).
41 Stephen Yarbrough, After Rhetoric, p. 172.
42 see Quine, ‘Ontological Relativity’.
43 Davidson, ‘A Coherence Theory of Truth and Knowledge’, p. 318.
44 Davidson, ‘Communication and Convention,’
45 Giovanna Borradori, ed., The American Philoso- pher: Conversations with Quine, Davidson, Putnam, Nozick, Danto, Rorty, Cavell, MacIntyre, and Kuhn (Chi- cago: University of Chicago Press, 1994), p. 41.
46 see Rorty, ‘Wittgenstein, Heidegger and the Reification of Language’, in Objectivity, Relativism and Truth, passim.
47 Yarbrough, p. 132.
48 G. A. Olson, ed., Philosophy, Rhetoric, Literary Criticism: (Inter)views (Carbondale: Southern Illinois Press, 1994), p. 20.
49 Patterson, Realist Christian Theology, p. 43.
50 Rorty, Philosophy, p. 265.
51 In Rorty’s sense of not giving an answer to the sceptic. cf. Davidson, Reality Without Reference, in In- quiries, passim.
52 Ronald Thiemann, Revelation and Theology: The Gospel as Narrated Promise (Notre Dame: University of Notre Dame Press, 1984).
53 Reinhard Hütter, Suffering Divine Things: Theology as Church Practice (Grand Rapids and Cambridge:
Eerdmans, 2000), p. 48.
54 Richard Rorty, Objectivity, Relativism and Truth, p. 152.
55 Kevin Schilbrack, ‘The Study of Religious Belief Af- ter Donald Davidson’, Method and Theory in the Study of Religion 14/4 (2002), pp. 334-349, p. 7.
56 See Mike Higton, ‘Frei’s Christology and
Lindbeck’s Cultural Linguistic Theory’, Scottish Journal of Theology, 50 (1997), 83-95.
57 Hans Frei, Types of Christian Theology, edited by George Hunsinger and William C. Placher (New Haven:
Yale University Press, 1992), p. 42 (my emphasis).
58 Davidson, ‘A Nice Derangement of Epitaphs’, p.