Alright: we've had Bill's reflection on the first chapter of Rowan Williams' 'The Edge of Words'. We've had my (unassailable) defence of the idea that the philosophy of Wittgenstein argues against the possibility of 'traditional' natural theology, then Bill's (clearly flawed) counter-argument. Now it's time to belatedly turn to Chapter 2: Can We Say What We Like? Language, Freedom, and Determinism.
What is The Bushy Eyebrowed One Trying to Do Here, and How?
So- in Chapter One (as I understood it) Rowan Williams suggested that be might be able to pursue a kind of natural theology by analysing the nature of language. He tries to carefully couch this suggestion in an attitude which is, prima facie, hostile to natural theology as traditionally understood. He tries instead to point towards how navigating to the edge of words might show the outer limits of intrinsic human possibility, and so the point at which God has to enter into speech, life, and everything.
Chapter 2 is the first part of this analysis of language as a natural phenomenon. The chapter is devoted to the question of what it means for speech to be free- that is to say, for meaning to not be entirely determined by the causal relations which it is the business of natural science to explore and express, such that it is instead a somewhat anomalous function of the interaction of free beings with the world.
As far as I can see, he does this by attempting to break down a distinction implicit in Richard Rorty's late writing- the distinction between an mechanistic, 'meaningless', and absolutely determined material reality and the 'meaningful' noises we arbitrarily use to designate them. He attempts to do this in a fashion which might seem somewhat roundabout, but allows him to (surreptitiously?) bring the question of theodicy into his analysis of the freedom of language. I'm going to try and briefly describe his overall movement in breaking down the distinction between reality and language, then briefly describe how the analysis of the ethics of truth telling through which he does this brings the question of theodicy into his writing. I'll then end by posing a few questions for reflection, relating back to some earlier posts on this very blog.
Attacking Rorty's Senseless Distinction
Williams begins with the basic assumption of freedom, employing what I take to be a fairly Humean/Wittgensteinian anti-sceptical argument (one which at the time presupposed a certain amount of speech-act theory), which holds that any denial of freedom by asserting an absolute determinism is a performative contradiction. In his words:
'To give reasons for believing determinism is true is to undermine determinism. To articulate the evidence is to relativise it, because to assume that the noises I make in defending determinism have the property of causing you to believe it is manifestly unfounded, and dangerously near to being a flat contradiction of the warming not to assume that a state of belief can be caused by anything except a set of immediate physical causes.' (p36)
From the point of view that absolute semantic determinism is impossible to argue for on pain of performative contradiction, Williams looks at one recent attempt to explicate the freedom of language in the context of a predominantly determinist world-view: namely, that of Richard Rorty, who 'argued a strong case for distinguishing between the… determinist account we have to give a physical causality and the radically underdetermined possibilities of meaning' (p37). Under this picture physical causality is indeed utterly deterministic, but the words we use more or less entirely arbitrary. We thus have two almost entirely independent strata- that of reality and that of speech. Speech is free precisely because it is distinguished from reality in this way, and so is not determined by the same constraints.
Williams, employing the the thoughts developed in Roy Bhakshar's 'Philosophy and Freedom', argues that this analysis in unsatisfactory (it is also, if I my say so myself, guilty of the same kind of grandstanding attacked by Alec in his post on Milbank). He points out that Rorty's account 'seems to be assuming that we live within a set of determined constraints that cannot be materially affected by our liberty to say what we choose' (40), but that this assumption is the very thing ruled out by the impossibility of attempting to defend determinism argued above. Contra Rorty, 'a closed-system determinism is impossible to sustain; we cannot avoid assuming that we are responsible for at least some of our physical actions, and our linguistic behaviour is one of the prime candidates for such responsible behaviour.' (40)
What Rorty has missed is that speaking (and so, we can infer, language too) is itself 'a material phenomenon' (35). Taking another Wittgensteinian line, Williams observes that 'we shall always have learned how to identify and describe' (41). Language always arises out of a material relationship between world and speakers, a relationship which is essentially conditioned by the context within which those speakers are taught to speak. As such, our speech is always in some real sense constrained by that reality, such that 'the idea that we are free to give what account of 'reality' we choose, that there is no difference between 'telling a story' and 'reflecting the world' fails to deal with the actual constraints that speech works with.' (41)
And yet: still 'we know that what we say is not actually dictated by what is simply there.' (42) So, what are we to do? The answer, painstakingly and obliquely developed through an analysis of what it is to lie, and so what it is tell the truth, is that 'the connectedness of language to what is not language is a shifting pattern of correlation, not an index-like relation of cause and effect.' (60) Undergirding this view is that idea that there is not in fact a sharp cleavage between deterministic reality and undetermined speech- rather, a world which is 'irreducibly charged with intelligibility' (64) underdetermines the precise form and matter of the language which we learn to use as its inhabitants. The freedom of language is thus premised upon the idea that the necessary postulation of semantic indeterminacy undermines the idea that world itself utterly determines everything within it- the very idea which rendered the freedom of language a problem in the first place.
The fact that the causal relation between world and meaning (which is through and through a natural, material relation) is underdetermined leads to a practical conclusion. Because there is no absolute causal relation between state of affairs X and the linguistic utterance 'X', 'we have an indeterminate horizon [at both ends]: the object is not to be exhausted by the multiple representations that keep being generated; the subject is not to be constrained by the limits of description or reflection.' (61)
This idea in turn informs our view of what language actually is, and so forms the first strut of Williams' attempt at a natural theology grounded in an analysis of language. To let Williams speak for himself;
If we do indeed have to come to terms with the awareness of our location, we also have to reckon with the fact that the way we learn, and represent what we have learned, is by working with various schemata which do not map neatly on to each other, and by projecting structures of representation which allow diverse aspects of what is encountered to emerge into view. We learn and represent our learning, in other words, in a mode that (i) assumes we have no final perspective on what is encountered (there is always more to say) and (ii) assumes that what we say alters what we can say next, so that the deposit of one essay in representation affects and (hopefully) enlarges what we go on to say.' (61-2)
This leads to Williams pointing forward, saying that 'in the context of the overall argument of this book, what is important is that the 'freedom' of language requires an anthropology, a picture of what is distinctively human in terms of receptiveness to a set of signals from the environment that do not allow of a final and ideally complete reading.' (62)
This is the first major fleshing out of what I take to be the central contention of Williams' book: that human language (and so the human will itself) is both dependent and underdetermined- it depends for its meaning upon a world to which it must respond and a social context which shapes the particular form of that response, but can still always be responded to differently. This indeterminacy is itself, meanwhile, a way of explicating human freedom in linguistic terms- the underdetermination of meaning necessitates the fact of freedom, whilst the fact that meaning still arises from an encounter with- and so is still responsible to- reality means that this freedom is not absolute: the fact that we are free does not mean we can just say whatever we want.
Misrepresentation and Theodicy
This, I believe, is the movement Williams charts in this chapter. We must still give an in depth account, however, of how precisely he grounds the conclusion that language is 'not an index-like relation of cause and effect.' I hope I can be forgiven for putting the actual mode of reasoning second- my reasoning in this decision being that it is well worth taking the time to develop the relevance of his reasoning to a linguistic theodicy outside of the overall movement of the chapter. So, let's do that now.
Williams begins with the work of George Steiner in 'After Babel: Aspects of Language.' His thought hinges upon a lengthy quote;
'I believe' [Steiner] writes, 'that the question of the nature and history of falsity is of crucial importance to an understanding of language and culture. Falsity is not, except in the most formal or internally systematic sense, a mere miscorrespondence with fact. It is itself an active, creative agent. The human capacity to utter falsehood, to lie, to negate what is the case, stands at the heart of speech and of the reciprocities between words and world… We are a mammal who can bear false witness.' (45)
Condensing the argument somewhat, William reasons from this as follows- a lie is not just a mere miscorrespondence with fact, but a function of an innate human capacity. The fact that this is a significant part of the character of a lie suggests that the character of telling the truth is itself complicated. And after working through a particular kind of Augustinian view of truth-telling, Williams notes that the possibility of deceit is intrinsically tied to the possibility of error and misrepresentation. It is possible for us to deliberately misrepresent what we see, but we also, all of us, do so accidentally as well: 'any serious interpersonal exchange involves moments when we struggle for words; when emotion of one kind or another leaves us baffled and inarticulate; when we cannot without a sense of dishonesty reproduce what we have said or heard in other circumstances apparently similar.' (57)
It is this possibility of misrepresentation, which arises out of reflections on both the nature of the will and the limitations of our linguistic capacities, which leads to the breaking down of a deterministic causal relationship between word and reality. The causal relationships in physical reality are grounded in their consistency- when we drop a pencil, it always falls. This is not the case with language: when we have a particular experience and try to communicate it, there is always an active possibility of misrepresentation (even and- if we really listen to Wittgenstein- especially when the words and experiences are familiar to us). And it is the breakdown of this posited causal chain which grounds the claim that;
'The form of an utterance cannot in itself tell us what is the case; the relation between the noises we make and any states of affairs in our environment is anything but straightforward, and the very idea of representation, or even a description, of states of affairs by ways of noises highlights the point about fundamental unlikenesses between what is represented and the symbolic medium of representation.' (53)
This is where we can continue back into the movement of breaking down Rorty's distinction above, if we want to. Let us instead, however, attempt to draw out the depth of how the possibility of error in Williams' analysis grounds a potential theodicy. First, let us note that the possibility of misrepresentation is not, here, something which comes after the Fall. Yes, 'Speech can betray… But it seems we must say that language is, from the beginning [my emphasis], capable of such use, for the simple reason that it presents us with choices about how we make representation- even how we 'describe', in fact.' (51) And though we are always faced with 'the temptation to seek to reverse this turning away as if it were the original sin of language, and to try to restore an unmediated reflection of what is there, a sort of unfallen descriptive clarity…, this must be resisted.' (54)
Thus we arrive at both an (implicit?) anthropology tied into account of linguistic freedom: 'our sense of what is distinctively human is bound up with our ability to be wrong or even untruthful in our representing of the environment.' (60) And here, at least, this sense of what is distinctively human goes right back to Eden.
This is very consistent with, in particular, Bonhoeffer's account of the Fall in Creation and Fall- if I remember correctly. There, the Fall is brought about by the fact that the words of the serpent can be misunderstood, and so misrepresented, by Adam and Eve. What we have here, then, is an anthropology which includes within it the possibility of falling from grace- and so an account of the possibility of evil.
According to Williams, however, this possibility of misrepresentation is what grounds both moral responsibility and human fellowship. On page 59, he writes that 'we are looking at practises of handling frustration and bafflement. And yet to speak in these terms about speech is also to affirm the freedom of our speaking in a way that is both metaphysically and morally significant'; and on page 58, 'what brings me closest to my fellow-humans is the fact that we are all incapable of making ourselves transparent, all faced with the limits of bodiliness and mortality and the diversity of our histories and everything else'. (58) Just like a classical theodicy, then, the importance of this freedom to err is tied to the fact that the possibility of error makes us our actions morally significant- and so our love of God and neighbour genuine love. In this account of language, then, we are faced with a possible way of expressing the necessity of the Fall on the basis of the necessity of error, that error being necessary to ground the fact of our freedom to love or not to love (without which we could not love at all).
I'm going to end with a few thoughts of my own, before posing some general questions.
First of all, I am sympathetic to this line of reasoning. I think it does a good job of connecting up language, life, and the world in a way that does full justice to the complexity of each. I have not dwelt upon (or mentioned at all) Williams' notion of symbolic representation (largely because it seems to me inessential to the basic argument), but I also think it approximates to the truth.
I have only three comments- the first is that I am not entirely sure the reasoning isn't circular in some way: in order to arrive at his two horizons of indeterminacy, Williams has to assume a prior indeterminacy, which he locates in the possibility of misrepresentation. It is, however, possible to understand misrepresentation in a way which doesn't require indeterminacy, both scientifically and religiously (I'm thinking of a strongest possible form of predestination). I don't think we have to agree with Steiner's account of falsity, and if we do, I don't think we have to draw the conclusions from it which I think Williams does. As it happens, I do agree with Steiner (with the support of Saturday Morning Breakfast Cereal), and I also agree with Williams, but I think he could be more explicit about the possibility that his line of reasoning, in at least some sense, assumes its own conclusion. I don't think this is fatal to his argument, of course- but it could be stated more clearly. (It's also entirely possible that he has stated this clearly, but that I missed it as I was working through my notes to write this post up).
The second is a question: where is Derrida? The line of reasoning pursued here is about as Derriderian as possible, which even a cursory reading of the Limited Inc. part of 'Limited Inc.' would show. It works in a structurally near-identical way, using the possibility of error to break down the purity of the logical concepts which ground causally deterministic world-views, and locates that possibility as essential in the same way Williams does. Nor is Williams totally unaware of Derrida's writing, as a citation on p152 demonstrates. Perhaps he hadn't read Limited Inc, or perhaps he thought it might make people switch off to give the name a prominent place, but either way, it seems a missed opportunity to expand the theological importance of a thinker viewed with suspicion by many of the people likely to read published Gifford Lectures.
The third is that the account of humanity here could lead to some interesting thoughts about the freedom of animals. I recently learnt that squirrels can lie- they will pretend to bury nuts when they know other squirrels are watching, but will in fact hide the nut elsewhere. If a squirrel can do this, and if we ground our understanding of human freedom in terms of the indeterminacy entailed by the possibility of misrepresentation, then I think we could open a very interesting route to expanding our understanding of non-human animal consciousness.
(A fourth, actually: Williams' line of reasoning connects him to Barth's evaluation of Thomas Aquinas, where Barth wrote in a letter to Thurneysen saying that the Subtle Doctor 'knew everything, but everything, leaving aside the one thing he didn't know, viz. that man is a liar.' (Barth and Aquinas: An Unofficial Catholic-Protestant Dialogue, p99) Not going to flesh that out now, but it seems worth mentioning!)
And the end- the basic questions at which are as follows:
1. Does any of Williams' thought make sense?
2. Have I clearly and accurately communicated what he is trying to say and how?
3. How does this relate to other theological and philosophical accounts of human freedom, in the context of ultimate dependence?
4. How does it relate to other areas of philosophical and theological inquiry?
On a more general level, I would be interested to develop how what Williams has written here relates to a discussion between Emily and myself on these pages (the next response to which I am still working on!) about the Problem of Evil. Specifically- does the fact that we are created with a capacity for misrepresentation suggest an evil or incompetent creator, if we allow that freedom does not entail error (even if error entails freedom)?
And on a far more pre-emptive note- Williams is going to, later on, say that it is constitutive of humanity that there is always the possibility for 'going on', in virtue of the indeterminacy, and so impossibility, of final expression. If we read this in terms of Wittgenstein's sense of the goal of philosophy, namely, to be able to stop asking questions every now and again, how are we to understand this in relation to the nature of human suffering and our ways of trying to ease it? This question might not make much sense on its own, but it's a thought I had often whilst reading the book, and I think it is prefigured in the discussions here.