“… the world is all outside: it has no inside.” — Ralph Waldo Emerson, “Experience”

1. Soft Behaviorism

The atmosphere of Oxbridge in the 1940s and 50s was one in which the rejection of various forms of idealism was in tension with new efforts to resuscitate inner experience. That inner experience required resuscitation was due in part to the far reaching influence of Gilbert Ryle’s attack on the “absurd” picture in which every observed action is imagined as accompanied by a shadow process hidden away in a separate place called the Mind, which he called “the dogma of the Ghost in the Machine.”1 And Ryle was himself indebted to what was by far the most pervasive style of thinking in Oxford at the time, the linguistic analysis (later “ordinary language” philosophy) that followed from lectures Ludwig Wittgenstein had been giving in Cambridge in the 1930s, up through his posthumously published Philosophical Investigations (1953). The latter work succinctly encapsulated a position that seemed to echo Ryle’s: “an ‘inner’ process stands in need of outward criteria.”2

Already in her first book Iris Murdoch voiced skepticism about this picture. “The world [described in Ryle’s] The Concept of Mind,” she wrote, “is one in which people play cricket, cook cakes, make simple decisions, remember their childhood and go to the circus; not the world in which they commit sins, fall in love, say prayers or join the Communist Party.”3 The oft-quoted bon mot neatly captures what she found unpersuasive in the critique of the dogma of the ghost in the machine: the first set of activities seem amenable to what can be observed in behavior (decisions are “simple” insofar as they issue in discrete actions) alongside a kind of caricature of inwardness (ah, let us recall the innocent days of childhood); while the second seems to evade outward verification. To commit a sin, fall in a love, say a prayer, or come to the realization that one would like to become a member of the Communist party are not occurrences that seem straightforwardly readable off external behavior. They seem to be things which occur privately.

Of course, Murdoch’s list is reductive, and there is more going on in Ryle than her tidily dismissive taxonomy. Ryle is not a “behaviorist” in the way, for example, Vienna Circle positivism thought all statements not verifiable by observation were meaningless.4 Ryle’s critique of the ghost is rather a “soft” behaviorism in which, as Julia Tanney puts it, “statements containing mental terms can be translated, without loss of meaning, into subjunctive conditionals about what the individual will do in various circumstances.”5 The grammatical point would seem to leave room for inner episodes just insofar as talk of such states have public rules (outer criteria) for their application. To get soft behaviorism into view, Tanney runs a thought experiment about an extra-terrestrial anthropologist investigating two soccer kicks. In the first a player kicks a ball into a goal, and nothing of any special significance follows. However, a second kick—in which “[t]he trajectory of the ball, the style of the kick, [and] the angle it makes” are identical to the first—results in wild cheering, a flurry of complex financial transactions, and celebrations in the street (we might imagine the first kick to be a player practicing alone, while the second is the winning goal in a world cup match). The anthropologist is perplexed about why such vastly different consequences follow from two identical physical operations, and so accordingly collects the two sets of balls, goals, nets, accurate replicas of the anatomical structure of the players’ legs and feet and begins earnestly studying each in search of the difference that could explain the divergent results. Tanney says the anthropologist is “looking in the wrong place for an explanation,” and by analogy the same may be said for the category mistake of the ghost in the machine: by looking “inside” for the referent of the meaning (or intention) of some observable action, one forgets the complex outer structure of concepts.6

While much has since been written about the relation of mental happenings to outward criteria, that work tends to follow the problem into various kinds of skepticism.7 Here I want to look at Murdoch’s differently attuned understanding of how inner experience is compatible with Ryle on the ghost and Wittgenstein on public criteria, as well as her occasional interlocutor Elizabeth Anscombe’s account of the relation of intention and action and, in a last section, Hugh Kenner’s elaboration of what T.S. Eliot called the “objective correlative.” In such examples the outer (observable) structure of concepts doesn’t so much block or occlude access to the inner as invite us to consider what it would meant to think that experience, intention, emotion—to use the words of Murdoch, Anscombe, and Eliot respectively—have an outside structure. The complex outer structure of concepts imagined in the Rylean-Wittgensteinian pictures doesn’t commit one to thinking that the inner doesn’t exist (or to mere behaviorism) but rather that experience, intention and emotion might be grasped without privileged access to something hidden or private.

2. Rings and Ranges

In an early piece, “Thinking and Language” (1951), Murdoch may be heard as at once inheriting, deepening, and challenging the myth of the “inner” as linguistic philosophy had imagined (that is, denigrated) it. “I have used the word ‘experience’ … which we have been advised against doing,” she notes, so advised presumably because of the murkiness of the datum to which talk of the mental might apply, whereas “mental terms do have clear and determinate conventions of use in connections with modes of overt behavior.”8 Murdoch’s explicit concern here with “use” invites an ordinary language sort of analysis: just what are we doing when we use the word experience on some particular occasion? One kind of use might be of the sort found when applying for job and noting that one has “five years’ experience” (say, hands-on experience operating a forklift). Another might be experience in the sense of qualia: of noting what it is like to taste coffee or hear a whistle or of skiing, perhaps with the implication that no verbal description quite captures the composite feeling of gravity pulling against your skis, the snow hitting your cheeks, the view of the mountains receding into the distance, etc., i.e., you “just have to experience it.”

Murdoch seems to blur these two kinds of use in a paper a few years later, in which she modifies the question of the inner, asking, “In a philosophical analysis of morality, what place should be given to the ‘inner life’?”9 In this case, the constraint of limiting talk of inner experience to what can be captured in outward criteria begins to seem unsatisfying. According to the dominant view:

the material which the philosopher is to work on is simply (under the heading of behaviour) acts and choices, and (under the heading of language) choice-guiding words together with the arguments which display the descriptive meaning of these words. Here two philosophical conceptions reach out towards each other and, in a hazy region, seem to meet. On the one hand there is no inner life, and moral concepts too must have meaning through definite external criteria. On the other hand, morality is a choice, and moral language guides choice through factual specification. The result is a picture which seems to have the authority of the modern view of the mind …. (VCM)10

If Murdoch wants more than a “hazy region” where acts meet up with choice-guiding words—exasperation with the limits of “the “modern view of the mind” (i.e., linguistic philosophy) —this is going to require a different way of understanding the nature of concepts. “A moral concept seems less like a movable and extensible ring laid down to cover a certain area of fact and more like a total difference,” Murdoch writes, and so “[w]e differ not only because we select different objects out of the same world but because we see different worlds [such that] differences of moral vocabulary betoken[] different ranges and ramifications” (VCM, 82). By “range” I take it she means something like an entangled, reticulated network, while “world” seems relatedly to be a whole surveyable geography of ramifying concepts.

The move from concepts as discrete rings to concepts as reticulated ranges does not however entail “abandoning the linguistic method [of Ryle and Wittgenstein, but] rather implies taking it seriously” (VCM, 84). Indeed, something very like this move comes through at moments in Wittgenstein, as when he imagines how our words (concepts) are not “clear-cut” [Scharfgeschnittenes] such that when we attempt to analyze them we feel “as if we had to repair a torn spider’s web with our fingers” (PI, 106). While the tone in this section of the Investigations seems to caution against getting entangled in hopeless complexities comparable to the manual repair of a spider’s web, it also captures what Murdoch means by the difference between discrete cutouts and woven, reticulated networks. Near the end of “Vision and Choice” Murdoch makes explicit her debt to Wittgenstein on this point, noting how differences in “moral pictures” should be understood along the lines of Wittgenstein’s claim that “What has to be accepted [as given] is … forms of life (Lebensforms)” (PI, 226 [VCM, 97]). The implication here is that a range of reciprocally entailing concepts constitutes a form of life.

Murdoch’s longer, more ambitious essay, “The Idea of Perfection” (1964), returns to the relation of decision and action and again concedes that the inner must be understood through public criteria. I learn the concept of “decision,” she notes, by watching someone who says, “I have decided and who then acts,” such that “[t]he concept has no further inner structure; it is its outer structure.”11 The view is something which, as Murdoch says, “originates in an argument in Wittgenstein.” Though she doesn’t elaborate the point, the argument she refers to is likely the following thought experiment:

Suppose that everyone had a box with something in it which we call a “beetle”. No one can ever look into anyone else’s box, and everyone says he knows what a beetle is only by looking at his beetle. — Here it would be quite possible for everyone to have something different in his box … But what if these people’s word “beetle” had a use [Gebrauch] nonetheless? — If so it would not be the name of a thing. The thing in the box doesn’t belong to the language-game at all; not even as a Something: for the box might even be empty. — No, one can ‘divide through’ by the thing in the box; it cancels out, whatever it is. (PI, 293)12

Whatever private introspection accompanies the use of “beetle,” its presence or absence floats free of the role the word plays in the language. Like Ryle’s ghost, the inner thing to which “beetle” is presumed to refer is a myth (or at best irrelevant). It is rather outward criteria that give “beetle” its meaning for the community. Here Murdoch again notes that in the face of “hopelessly hazy inner phenomena” we ought to accept our reliance on “the ‘outer’ face of the concept, since the inner one is so vague” (TIP, 12). But then she immediately pivots to what she finds unsatisfying in the idea that “[w]hat I am doing or being is not something private and personal, but is imposed upon me in the sense of being identifiable only via public concepts … Reasons are public reasons, rules are public rules” (TIP, 15).

If there is reluctant acceptance of the Rylean-Wittgensteinian picture here, it is one which remains unsatisfied with what might be taken to be some form of behaviorism (soft or otherwise).13 The picture of concept application not as decision-guidance but as a range of use which traces in relief whole forms of life should presumably include ways of referring to what Murdoch elsewhere calls “lots and lots of objects, … in orbit in as it were in inner space” (TIP, 13). If a concept’s structure is on the outside, how are we to square this with the idea of inner happenings that remain out of the range of observation? To begin to get an answer here we need to look at one of Murdoch’s own thought experiments.

3. M. and D. (and S.)

A mother-in-law, M., believes her daughter-in-law, D., is “unpolished and lacking in dignity and refinement,” is “insufficiently ceremonious,” and sometimes brusque to the point of rudeness. From such assessments M. concludes that she is a “silly vulgar girl” and that her son has married beneath him (TIP, 18). M. arrives at her view of D. as a function of a conceptual scheme which Murdoch says “imprisons” M. To be imprisoned in this way is to draw reflexively upon an unexamined range of concepts. There is simply no other conclusion to be reached about D. given D.’s manners, dress, and accent (accent here presumably a British class marker and not a Japanese or Israeli or German or Italian accent). The scene setting of the thought experiment is necessarily spare on this point: given the emotionally complicated scenario of newly acquainted in-laws, we are presented with a controlled picture of a schema of appraisal.

Presumably drilled since childhood in rigorous patterns of etiquette—nothing if not public criteria—M. may conceal her low view of D.; indeed, D. may never know of M.’s disapproval, may be living abroad, may even be deceased. But at some point, M. may say something along the lines of “Let me look again.” The further effort of “looking” here, extending a more attentive and caring notice of D., seems to require an idea of inner experience as something that may undergo revision and reassessment. M. inwardly reviews her experience of D. and comes to see and evaluate her differently. She may consider the possibility that she is herself stuffy or a snob or that she is jealous of her son’s attachment to his new wife. Remembering Murdoch’s concern to hang on to a linguistic analysis of the sort she associates with “a certain argument in Wittgenstein” we might say M.’s revaluation of D. is nothing more or less than the revision of a network (or spiderweb) of public concepts, and so to a whole new range use: D. is no longer “vulgar” but “refreshingly simple”; no longer “insufficiently ceremonious” but “spontaneous”; not “bumptious” but “gay”; not “tiresomely juvenile” but “delightfully youthful” (TIP, 18). Such revision would not be the sort in which M. “merely changes the application of an unrevised set of concepts.”14 What happens “inside” M. is not to have the thought: I had previously found D. to be insufficiently ceremonious, but I’ve come to see that she is sufficiently ceremonious, and so on. Rather, M.’s coming to see the “unsatisfactoriness of whole ranges of concepts she has earlier employed unreflectingly” leads to a new evaluative Lebensform and so a change of world.15 We might imagine further stretching the range of use upon occurrences “in” M. that further refine her sense of D. (such as: “sure she has a certain candor and charm, but it is liable to spill over into an insouciance not always appropriate for the occasion” or some such further revision).

Without wishing too quickly to assert an untroubled affinity between Murdoch on conceptual range and Wittgenstein on outward criteria, we may wonder whether M. can be seen as falling into the confusion Wittgenstein’s “S.” man falls into. The S. man is involved in an eccentric ceremony by which he christens a rule for marking the occurrence of an inner happening by writing “S.” in a diary. “What reason have we for calling ‘S’ the sign for a sensation?” asks a voice, “For ‘sensation’ is a word of our common language, which is not a language intelligible only to me” (PI, §258). That the example is a prolongation of the “beetle” scenario becomes clear when the voice notes that “it would not help either to say that it need not be a sensation; that when he writes S. he has a Something …  But ‘has’ and ‘something’ also belong to our common language” (PI, §260). Then comes the diagnosis: the S. man “must inwardly resolve to use the word in such-and-such a way. And how does he resolve that? Should I assume that he invents the technique of applying the word” (PI, §262)? The problem here is that such a technique could not be arrived at through “mental pointing” but must in the end stand in need of public criteria.16 If both the Beetle people and the S. man are tangled up in a confusion about private introspectabilia, isn’t this comparable to what M. does when she “privately” revises her moral judgments about D. (in D’s absence)? Doesn’t M. “inwardly resolve” to use her new terms of appraisal and so “invent a technique” for how to use the new words?17

4. Beyond Physics

If Murdoch had been working through an argument “that originates … in Wittgenstein” about inner processes and outward criteria, Elizabeth Anscombe considers the same problem, but with the focus on what counts as an action. Early in her 1957 book, Intention, Anscombe distinguishes between the concepts of intention and prediction. Saying “I am going to fail this exam” might be an effort at prophesy or a statement of intent. If you feel you haven’t studied enough, you could say the sentence is a prediction. But if you take great care to get every last question wrong on, say, a macro-economics exam so as later to display the paper on which the failed exam in printed in a glass cube hung from the ceiling of the MoMA (call the piece Failed Economics), you may be said to have intended to fail the exam. Other moments in Intention involve observing an action without any overt statement of intent (or prophesy). Anscombe gives the example of sitting in a chair writing and imagining an observer asserting that her intention in that moment is to write. It seems unlikely, she notes, that the observer would presume her activity of putting pencil to paper while seated as an effort at “affecting the acoustic properties of the room.”18 So, action and intention complement one another for Anscombe in a way that seems expressive of the fit between inner process and outward criteria.

But is there an “inner process” here at all? A version of that question is taken up in one of Anscombe’s thought experiments which involves a wartime scenario in which certain actions will lead certain people to die.19 Imagine that a man is pumping water. This action supplies water to the inhabitants of a house. Another man adds lethal toxins to the water as it flows into the house. Inside the house are Nazis plotting genocide and whose death will be good for the world. The questions for Anscombe are how we are to understand what the man does and why he does it. For not only are pumping and poisoning different actions, so are moving an arm up and down, replenishing a water supply and saving the Jews.20 That the action splits into (at least) four answers to the “why is the man pumping?” question—to pump water, to supply water to the house, to introduce toxins into the bloodstream of the inhabitants, to save the Jews—is all part of the analysis of the concept of “intention.” The upshot is that there may be no need to posit some inner process as the source of the action, since to know how pumping fits together with supplying water and how water may be contaminated and how contaminated water will affect the physiology of the one who drinks it and so on is to know one’s way around in a form of life. This is the same as saying that the concept of “intention” has an outside structure. Its meaning may be given by noticing what is being done by whom in what ways at what times through the observation of how actions ramify.21

In a recent exchange about Anscombe’s account of action Walter Benn Michaels notes a distinction which maps fairly closely onto what I’ve been calling the “inward process and outward criteria” question: Anscombe’s concern with “what someone does” (what can be observed) and the literary theoretical question of the internal “state of the author’s mind.”22 Michaels is making a point about what Anscombe calls “going beyond physics”: “What’s suggested by the juxtaposition of starting with what physically takes place and going beyond physics is the idea that what physically takes place can’t be reduced to physics. If what physically takes place is intentional, and what happens is what physically takes place, what happens must be already under the sign of the intended” (EWS). I take Michaels here to mean that to recognize intentionality is to identify something not directly registered in observation (verification, measurement). But that doesn’t mean that we don’t see intention. While Michaels here refers to a different example in Anscombe’s book (I, §45), the description will also work with the pumping example. We may get answers, under different descriptions, to versions of the “why does he pump?” question—supplying, poisoning, killing, saving, and so on. We can also find external (observable) answers to what he is doing but not intending—for example, as Anscombe puts it, he is not setting out to “generat[e] … substances in his nerve fibres” caused by the movement of his arm, nor is he intending to “cast[] a shadow on a rockery where at one place and from one position it produces a curious effect as if a face were looking out of the rockery” (I, 37). While the man does make the “face” appear when moving his arm, it seems right to say that this is not his intention, whereas other consequences of the action of pumping are part of the intention (provided he and the man adding toxins to the water are in cahoots). The point here is that he knows a certain way of going on in a form of life; the range covered by a conceptual cartography that instantiate a great many linked and overlapping normative practices. And this then seems a good way of explaining why, for Michaels, if you find yourself interpreting a piece of writing, you do so “under the sign of the intended.” If you are reading or interpreting something you are as it were observing a completed action—its having been composed.

 5. Objective Correlative

About exactly the time Anscombe’s Intention was published, Murdoch contributed a short piece to a Festschrift for T.S. Eliot’s 70^th birthday. While there is something of an irony to the historical turn backward here—the linguistic philosophy Murdoch and Anscombe wrestled with at Oxford stemmed in large part from a violent rejection of the British idealism Eliot wrote his dissertation on—the modernist literary theory Eliot created in incidental newspaper pieces and essays in little magazines between 1918 and 1920 form an uncanny analogue with the Rylean-Wittgensteinian problem that preoccupied both Murdoch and Anscombe.23 While Murdoch notes that Eliot “does not trespass on the field of technical philosophy,” she nevertheless sees his work as part of “shared concern with contemporary moral philosophers [namely] the idea that a destruction of morals is a destruction of concepts.”24

It would not be a stretch to say Murdoch found Eliot’s literary criticism to form part of what she had earlier called a “modern view of the mind,” as his claims about art and artists had by the time of her essay become something like received wisdom: that art is not the “expression of personality”; that the progress of an artist is one of “continual self-sacrifice, a continual extinction of personality”; that an artist is “constantly finding an object which shall be adequate to his feelings”; that art should “introduce precise emotion [by] bringing it back to the object”; and so on.25 While Murdoch does not mention it, all of her collected examples converge on Eliot’s enigmatic idea of the “objective correlative,” put forth in a 1919 review of a book about Hamlet:

The only way of expressing emotion in the form of art is by finding an “objective correlative”; in other words, a set of objects, a situation, a chain of events which shall be the formula of that particular emotion; such that when the external facts, which must terminate in sensory experience, are given, the emotion is immediately evoked.26

Eliot here imagines as an aesthetic problem the question I’ve been tracking: inner experience (some “particular emotion”) finds its correlative in an external form which “immediately evokes” it. While the words “formula” and “immediately” may seem exaggerated—suggesting as they do a natural transaction27—Eliot is mostly raising evaluative questions: What will best express a harmony of proportion between felt emotion and observable action? In the case of Hamlet, what the character says is out of sync with the unfolding of the action. Hamlet’s showy, ostentatious soliloquies are evidence of an “emotion … in excess of the facts as they appear” (H, 125). To get this relation right would be to get the inner happenings (emotions) to “terminate in sensory experience” in the right way. And to get that right would be to achieve an “‘inevitability’ [which] lies in this complete adequacy of the external to the emotion” (H, 125). Eliot’s aesthetic criteria are close to Anscombe on action: to get the right relation of dramatic action to emotion is to think that what someone does (or says) on stage just is the inner happening.

In his sensitive and acute 1959 study of Eliot’s poetry and criticism, Hugh Kenner notes that the “famous sentence about the objective correlative” emerges from a tension between “the action of the play” and Hamlet’s “anthology pieces dealing with the cosmos in general.”28 Kenner sees their asymmetry as “so little rooted in the particular action that an actor-manager feels free to shuffle them into coincidence with his notions of pace” (IP, 87). What goes wrong in Hamlet for Kenner’s Eliot is a failure “to fix … just that concatenation of incidents and images which would make Hamlet’s emotion comprehensible and inevitable, and so release the soliloquies just where they occur. … [Such a] concatenation is the condition of artistic ‘inevitability’” (IP, 102). Kenner’s remarks here suggest that getting the “release” point wrong (“freely shuffling” compositional elements arbitrarily) undermines the inevitability of an inner process—some motive, or feeling, or intention—terminating in some action. While Kenner in part attributes the generalizing ambition here to what he calls Eliot’s “dangerous gift of phrase” (“objective correlative” has a ring that made it amenable to institutionalization for a certain New Critical orthodoxy), that doesn’t diminish the genuine philosophical depth of Eliot’s example. The question Eliot seems to ask is: Would a work of art that got the objective correlative right dissolve the distinction between inner process and outward criteria?

Commenting on Kenner’s writing devoted to other examples from Eliot, Todd Cronan notes that in art, “actions are determined by their objects,” whereas in morals, “actions are ‘determined by their motives, which are hidden: hidden, often, from the actor.’”29 Let’s bracket for the moment whether we must always imagine morals deontologically—i.e., that motives are what count and not, say, consequences (the view Anscombe despised) or virtuous or unvirtuous character (the approach Anscombe ends up more or less adopting). Note how we’re back to the problem of what is hidden and what is observable: in art, action just is the work as it is made, observable in the form of a done piece. This comes through clearly in Cronan’s remarks on Anscombe’s pumping example: “The question raised by the pumping action is whether the pumper is intending to kill the Nazis in the house or was unaware of the consequences of the act of pumping water.” (I noted above that we ought to assume from the example that pumper and poisoner are in cahoots.) “[O]ne has to know where the action begins and ends, the action has to take on a ‘form.’”30

But when Kenner gets to Eliot’s later plays, the work of art “purposefully occludes,” as Cronan puts it, what someone really is doing. Let’s treat that thought as contrastive with how Kenner had described the “objective correlative.” What happens in the objective correlative, if things go right, is that what is felt on the inside is, to use Kenner’s language, “just that concatenation of incidents and images which would make [the character or dramatist’s] emotion comprehensible and inevitable.” But when Kenner turns to Eliot’s The Family Reunion, we find a play which “throw[s] attention onto the invisible drama of volition and vocation [such that t]he plot provides, almost playfully, external and stageable points of reference for this essentially interior drama” (IP, 288).31 This, Cronan notes, “‘throws attention’ back on the invisible act of creation.”32 It seems clear to me that Kenner, for whatever reason, deliberately inverts the objective correlative in his remarks on The Family Reunion. Instead of some “interior drama” issuing inevitably in an outward object, The Family Reunion throws attention back on something private and hidden. Why Kenner wished to make a symmetrical reversal of Eliot’s literary theory in his reading of The Family Reunion is an interesting question (perhaps this is part of his own playful intra-Eliot game of cross-reference in The Invisible Poet) not least because it stages, from both sides as it were, the problem of getting inner processes in alignment with outward criteria. The Eliot of “Hamlet and His Problems” wants total objectification of feeling in an exactly calibrated “release” point; the Eliot of The Family Reunion wants the reverse: a drama whose external apparatus is meant only to get us to pay attention to something invisible.

Both sides here chime with what Michaels calls “form that hangs on to process” (EWS). If the work of art manifests as outward form, the choices that went into its making—the whole hazy, tumultuous drama of an inner process, “lots and lots of objects in orbit in inner space,” as Murdoch put it—then this is not unlike delimiting the form of an action. For Kenner’s earlier Eliot, this is imagined as composition which objectively correlates to inner happenings; for Kenner’s later Eliot, it is a matter of composition that throws our attention toward the invisible nexus Cronan calls the “act of creation.”

In both of these scenarios “form hangs on to process,” in the sense in which “inner process” means authorial intention. But “form” here might also be heard in the sense of Lebensform. That is, as the whole ramifying network of overlapping conceptual agreements that make up our everyday practices and sense of going on; the form that tells you how pumping, poisoning, supplying, and saving go together; or how a shift in moral appraisal sets off a revision of a whole conceptual schema. None of this is to reject Wittgenstein’s “an ‘inner process’ stands in need of outward criteria” or Ryle’s critique of the category mistake of the ghost (PI, §580). It is to make a series of clarificatory analyses upon them.