Authors are reluctant to admit responsibility for the uncounted ways their work is misunderstood. They prefer to acknowledge the equally uncounted ways their work is understood, even if they are surprised by some interpretations they encounter. I am most grateful to the three participants of this symposium - John Gibson, Paul Guyer, and Mary Wiseman- for their work in understanding what I have written. They succeeded admirably well. In studying their contributions I became aware of their deep understanding of what I have tried to achieve. Interpretations grounded on understanding of the work which are at the same time surprising for its author lead to an important goal of the interpreting activity: the author's self-examination and self-discovery. I am most grateful to John Gibson for organizing both this symposium and its ancestor.

2. Reply to John Gibson

Without Gibson's careful reading of my book, it would have received some time ago its premature burial on the shelves of our libraries. His work in thinking and writing about it kept it alive; he knows it so well that at times I think of him as its coauthor, but this is impossible, for he found out about it only when the galley proofs became available. Still, when asked what I contributed to our field, I would suggest that even if we disagree on some issues, he has the best answers to this question.

Gibson was especially successful in catching the spirit of my insistence on the interpreter's voice, and in making sense of my critique of deep interpretation. I will discuss these topics at greater length in my replies to Paul Guyer and Mary Wiseman. Prior to these replies a remark will be useful.

In writing about deep interpretation, my aim was to defend the views of Marx, Nietzsche and Freud against their intellectual progeny. I had no quarrel with any extension of common sense views of interpretation provided by the great thinkers of deep interpretation. The focus of my objection was an accidental feature of these extensions in the hands of some modern day practitioners: the habit of disregarding significant and important properties of what required interpretation and transforming it into an interpreted object that was prefabricated to show the power of a given view of deep interpretation. In their hands, the extension of common sense views provided by the three great thinkers turned into ideologies that offered ready-made answers to all puzzling questions. My critical remarks were primarily directed against practitioners who could advertise for themselves under the heading, "Have theory, will travel."

In turning to problems in the practice of interpretation, Gibson would like to examine how my views can be applied in the interpretation of modernist art, especially poetry. A preliminary remark is in order. Do we have only one concept of interpretation that is applied in different fields? Or, do we have many interpretive concepts that are dependent on the topic that is being interpreted? In my writings I was guided by the hypothesis that we have only one concept of interpretation. If this is accepted, then we are committed to the claim that this concept is used not only across different fields and styles, but also within the same field at different times. Interpretive problems of modernist poetry are not radically different from classical poetry. This claim is controversial, and it will be rejected by all critics who admit that they are only sensitive to classical art, literature or music, but are blind or deaf to their modern or modernist forms. We are all partially insensible to some art forms, and we cannot expect that the words and gestures of other critics who have a less parochial view of their own fields will remove our insensibility.

Even amateur criticism requires that we use words and gestures in expressing our aesthetic delight. While I share Gibson's appreciation of Ashbery's poetry, I dare not speak about it. My prerequisite for talking about a poem is hearing it when it is read aloud, or sounding it out in a language that I can speak without a disturbing accent. In selecting lines of another modern poet and contrasting it with a line of a classicist poet, I expect to show how interpretive views can be applied across the same field at different times.

Jean Racine, Phèdre, (1677), I, 1, lines 34-36:

Cet heureux temps n'est plus. Tout a changé de face,
Depuis que sur ces bords les dieux ont envoyé
La fille de Minos et de Pasiphaé.

Verlaine, Chanson d'automne (1866)

Les sanglots longs
Des violons
De l'automne
Blessent mon coeur
D'une langueur

Line 36 of Racine's Phèdre has elicited more interpretive commentary than any other line in the French canon; Verlaine's six lines are a close second. The names of the unfamiliar mythical characters suggest that the reader consult the relevant entries in a dictionary of classical mythology. Yet, even if he knows that Minos was a reasonable judge in the underworld, and Pasiphaé was at times ruled by her passions, the content of that line does not explain why it was the target of critical commentary for more than three centuries. The content of that line is not more profound than Verlaine's six lines. So, what is it that draws attentions to these two fragments of language? We must focus on the sound structure of these language fragments if we wish to become aware of their impact on the listener.

To be sure, it is neither a necessary nor a sufficient condition of aesthetic delight that we talk about its object. We start talking about it when we are engaged in criticism or interpretation. In talking we may succeed or fail in providing ways of understanding that object and at the same time in offering reasons for its appreciation. In the case of both Racine's line and Verlaine's poem I would argue that the listener or reader ought not expect a profound content that waits to be revealed by a critic's ad hoc theory. This is also a controversial claim, and it will be rejected by critics who require for a critical judgment the investigation of content prior to the examination of the sound structure of a line or a poem. Regardless of the side of the controversy the reader adopts, after the initial steps we all proceed along parallel lines.

The procedures of critics who rely for interpretation on an examination of the content of these language fragments are far removed from my concerns with poetic language. These critics are surely in a better position for speaking about their own practices than others who do not share their views. Still, they would agree that within the limits of the views they defend, they provide the best available interpretation of what they were interpreting, they have satisfied the factual and normative constraints on interpreting, and all competent interpreters agree with their interpretations.

Other critics who focus on the sound structure of poetry rather than its paraphrased content refer to the contrasting sounds of the second (fille) and seventh word (Pasiphaé) in Racine's Phèdre, line 36. All but the first word of that line occurs in the characterization of Phèdre in the list of dramatis personae ("Phèdre, femme de Thésée, fille de Minos et de Pasiphaé"). Within this list, these words must be read for the biographical information they contain about Phèdre; when read for its content, this sequence of words is quite unremarkable. Yet if the same words occurring on line 36 are read aloud and we hear only the content of that line, then we are not listening to poetry. The poetic power of that line comes across to the listener only if he hears the contrast between the hard sound in 'fille' and the soft 'aé' sound at the end of the seventh word.

The contrasting sounds of the two words reveal the same information as an analysis of its content would have revealed. Focusing on her characterization in the list of dramatis personae, I read it as if it were Phèdre's calling card; focusing on line 36, I heard the contrasting sounds as a key to her complex character. We fail to appreciate Racine's poetic powers if we say only that both the characterization and line 36 foreshadow what will happen in this tragedy. Even authors of detective stories know how to leave hints about events that will unfold in their stories. And it is not enough to marvel at the great economy of means employed by Racine, who succeeded in telling us first in six and later in seven words what we must know to understand Phèdre's character. We start to appreciate poetry, and here I speak not only of Racine, when we realize the decisive role of the sound structure of language fragments in revealing what the content is about.

Similarly, when we focus on the sound structure of Verlaine's poem, we become aware of the long nasal vowels that convey gloom. When we listen to a classical, modern, or modernist poem and find the key to that poem, we become aware that most poems were written to be heard. Musicians can 'hear' music in their minds' ears, and they need not sound it out for criticizing a fragment of what they have 'heard.' In our practices of interpreting poetry, there may not be anything analogous to the musicians' experience. I would argue that most of us must rely on our biological ears for listening to poetry. By repeatedly listening to the same fragments of poetic language and sounding them out for our own pleasure, we come to learn by heart some lines of a poem. While we may remain silent about the cause of our aesthetic delight, if we wish to speak with others about our experiences, we must provide reasons for our interpretive decisions. What we say at this point must satisfy from our viewpoint the same constraints that must be satisfied by other critics who aim solely at understanding the content of these language fragments.

Interpreters may appeal to any extension of natural interpretation when they interpret poetry, including some that were created solely for the purpose of illuminating a line of a modernist poet. They may call this extension a theory with or without quotation marks, or they may not even give it a name. Many of these extensions may suggest to a given interpreter interesting remarks about a language fragment of traditional, modern, or modernist poetry. Still, after offering his remarks as an interpretation of that fragment, we are entitled to ask: did he succeed in illuminating that fragment or only in exemplifying an extension of natural interpretation? Now if we judge that the interpreter has succeeded in both tasks, his interpretation must be given a hearing and we must be prepared to accept his interpretation. To be sure, we may reject it for other reasons, but we cannot disqualify it because it is grounded on an extension of natural interpretation.

3. Reply to Paul Guyer

Paul Guyer describes my book "as a work on the conditions for the interpretation of acts in general and speech-acts in particular." The word "speech-acts" does not appear in this book. Of course, Guyer knows well why I don't speak about speech-acts. He spells out the reason in the second paragraph of his remarks: "Interpretation is needed only when the speaker's or agent's meaning is not immediately obvious, so Stern's account of interpretation is not meant to be a completely general account of understanding the intentions of others." Talking about speech-acts would be incompatible with my model of interpretation. As long as we are clear on this point, the critical reader must remain free to translate my words into his preferred idiom.

My understanding of interpretation will become clearer, if we focus on what Guyer calls my attack on deep interpretation, which is, in his judgment, the animus of my work. For reasons that will become evident, I must report on my perplexity upon first reading this claim. I could not understand on what grounds I stand accused of such a foul deed. I consider myself an admirer of the work of the great thinkers of deep interpretation: Marx, Nietzsche, and Freud. So, why would anyone want to accuse me of attacking deep interpretation?

Even for a moment I couldn't accept that this was an off-the-wall interpretation. (An example of such an interpretation is that Michelangelo's Man's Fall in the Sistine Chapel was created to serve as a landing surface for flying insects.) After all, Paul Guyer, the interpreter of my work happens to be one of the best philosophical interpreters of his generation. Still, given my strong resistance to his understanding, my first reaction was that his interpretation was a deep interpretation of my views. However, for me this was not a live option. For if I accept his interpretation as a deep interpretation, then at the same time I admit not only my own lack of sincerity but also open the door to questioning his sincerity. Why?

Authors may not be expert interpreters of their own books, but they are at least competent interpreters. Now as a critic of my book, Guyer expects that either all or some competent interpreters will agree with his judgment. Given that as a competent interpreter I disagree, he can expect only the agreement of some competent interpreters. Accordingly, he can no longer appeal to the Universalizability Principle; he can appeal only to the Restrictive Principle. (Of course, he may not want to appeal to any principle.) No doubt all interpreters who agree with him are competent. But it was primarily my resistance to his judgment that excludes from consideration the claim that all competent interpreters agree with him. As long as we recall the interpretive tradition from Shakespeare to Freud and beyond, no great acumen is required for raising here a question: isn't the strength of my resistance a mark of my insincerity?

At this stage my only choice was to label his interpretation an off-the-wall interpretation or a surface interpretation in the subjunctive mood. For the reason already mentioned, I have excluded the first choice. The second choice did not seem more promising. For if I admit that his interpretation is a natural interpretation in the subjunctive mood, then at the same time I concede that I would admit his interpretation, if I had all his information at my disposal, and I would be sincere and reasonable.

Let us pause here for a moment. Other details must be added before the reader of these lines reaches a judgment in this debate. Still, it can be seen that moral issues arise very quickly in interpretive disagreements. The details added will exemplify my views on the interpreting activity, and will enable the reader to reach a judgment on the extent of our agreement.

Of course, at any time while thinking about our interpretive disagreement, I could say that Guyer did not get me right, and this is all that needs to be said. It is easy to say this upon encountering an off-the-wall interpretation of my work or an interpretation that does not deserve serious discussion. But in the context of this discussion, it is important to offer at least a hypothesis: if he was mistaken, then why was he mistaken? Either one or both of us may be at fault. Let us first exclude some irrelevant interpretive alternatives.

Could it be that it is Guyer who is motivated by some unavowed interest in an alternative view of interpretation of which he is at least partially unaware? Note the reversal of our situation. Before I wanted to defend myself against his deep interpretation of my views, and now I offer a deep interpretation of his views. Both deep interpretations lack an essential element. Each of us would have to point to the moral failings of the other in support of the claim that the other is offering a deep interpretation. In political debates each side suspects the other's sincerity; of course, this suspicion is often justified. In debates about matters of religion, we often find ourselves in the position that we attribute insincerity to our opponents; no doubt, often we have good reasons for our views. But in our debate about deep interpretation there are no reasons for such suspicion. Hence, neither of us is free to argue that the other offers a deep interpretation. So, what are my choices in judging his claim that I have attacked deep interpretation?

Since I have excluded the other choices, I could admit that Guyer offered a natural interpretation in the indicative mood: all competent interpreters agree with his interpretation. Alternatively, I could admit that it is no longer an interpretation of my work that I attacked deep interpretation: it is a fact. To be sure, I will resist either alternative. Still, it was important to show that an interpretive disagreement can easily turn into a discussion about the facts of the matter. Interpretations are located between what I have called off-the-wall interpretations and facts. Within that space obviously false claims gain entry, but also misinterpretations, deep interpretations, adequate interpretations in the subjunctive or indicative mood, and even interpretations that are candidates of becoming at a later date obviously true claims. Deep interpretations play a very important role within that space; their defense against attack is at the center of my work.

Interpretive controversies are grounded on the possibility that at least one of the opponents is mistaken. From at least one participant's viewpoint, we reach a satisfactory resolution of the controversy if that participant comes to understand why he was mistaken or why his opponent is now mistaken. Failing such an outcome, each can only repeat that the other is mistaken. In many cases this is the best that can be done. Further inquiry into the reasons for a mistake that I have made or my opponent is now making lead very quickly to charges of deficient self-understanding or self-deception. Such charges always have a moral dimension. It is at this stage that the power and infirmity of deep interpretation can be evaluated. First, we must clear up a point that creates an obstacle to understanding.

According to Guyer, the real thrust of my "concept of deep interpretation is that it is one to which the interpretee is supposed to be, as it were, permanently resistant but in which the interpreter nevertheless claims to be justified." The word "permanent" occurs only once in my book in another context; "permanently" does not appear. If the permanent resistance of the interpretee is understood as a defining characteristic of deep interpretation, then my account is either false or incoherent.

Suppose I am trying to find a reason for someone else's mistake. Even suggestions that sound innocent can lead to deep interpretation. For example, I offer carelessness as a reason for the mistake. When the reason I suggested meets strong resistance and alternative explanations are deemed insufficient, it may occur to me that deficient self-understanding is a reason for the resistance. Of course, either one of us can be mistaken. But even if I am right, I would mention the deficient self-understanding of another person only if I am prepared to stand by the moral dimension of my charge. Charges of deficient self-understanding or self-deception always have a moral dimension. Bereft of such a dimension, there is no point to deep interpretation. Our great teachers of deep interpretation have taught us that every deep interpretation of the words and deeds of another person ends with the (often unspoken) concluding Rilkean insight: you must change...!

If the interpretee's permanent resistance were a defining characteristic of deep interpretation, then the call for change wouldn't make sense. Also, Freud, Marx, and Nietzsche would be reaching in their writings only the few-if there are any-who do not need deep interpretation. And this is false. It is part of the power of great theories of deep interpretation that they are addressed to all. Another part of their power derives from the fact that their interpretive methods have survived critical scrutiny. Their survival is secured by their critics as much as by their practitioners.

Praise for the power of deep interpretation must not hide an infirmity that comes to the fore only in a discussion between its practitioner and her interpretee. Contrary to Guyer, I hold that the practitioner often is justified in arguing that the interpretee "sincerely believes what he claims to believe but is blamably mistaken in so doing." Let us assume that the practitioner is not mistaken and that she knows better than the interpretee what he meant by his words and deeds. She relies on a set of interconnected beliefs grounded on an extension of natural or surface interpretation. By saying that the interpretee is blamable, we admit that although the interpretee was in a very good position to know what the practitioner attributes to him, the fact that he did not know it suggests his deficient self-knowledge. Call this deficiency self-deception or call it by any other name, and if you prefer, don't call it by any name. You happen to be in excellent company regardless of whether you affirm or deny the existence of the phenomenon of self-deception. (The Humean Kant and the Cartesian Sartre were on opposite sides on this issue.) Important is only that the interpreter attribute a moral failing to the interpretee.

Guyer suggests an alternative to my account. "But deep interpretation might also be thought of as explanations of why certain sorts of speakers have the intentions that they do, so that it is not false that the speakers have those intentions, but the fact of their having those intentions is to be explained in a way that the speakers do not realize or recognize. In this case, speakers would not need to be self-deceived about their own intentions but would rather be ignorant of the explanation of those intentions." Confronted with this alternative, I must ask: Can we attribute this interpretee's ignorance to a deficient self-knowledge and blame him for his moral failing? If the answer is affirmative, then this case is not different from other examples that call for deep interpretation; if it is negative, then there is no reason for deep interpretation.

Except in a political or religious context, we are seldom in the position of attributing deficient self-knowledge to another person while we are engaged in a debate with him about what he has said or done. Even if we are convinced that all competent interpreters would agree with our judgment of the interpretee while debating with the interpretee, and we are entitled to appeal to the Universalizability Principle, we restrict the scope of our judgment to some competent interpreters. It is in this context that I located the insincerity of deep interpretation: the deep interpreter relies on one principle privately and on another publicly. This is a small but indelible stain on our interpretive practices. In defending deep interpretation, we must accept that insincerity is its price on the rare occasions when we attribute deficient self-knowledge to another person.

How often did the great thinkers of deep interpretation engage their interpretees in a discussion about deficient self-knowledge? Hostile critics have argued that their views are generalizations from few cases. Even if the critics are right about the facts, e.g. about the very small number of patients Freud saw during his long career, this is irrelevant in an assessment of deep interpretation. For the power of deep interpretation reveals itself, not in the quarrels between practitioners and their critics, but in the work of interpretees in their process of self-discovery. Practitioners of deep interpretation are at their best when they rely on their interpretees to work on their self-interpretation. In optimal cases the self-interpretation will yield self-discovery.

Guyer charged me with attacking deep interpretation. I replied that it is the defense of deep interpretation that is at the center of my work. Who is right in this debate? Before the reader decides on this issue, caution is in order. At least four possibilities must be examined: one of us is right, both of us are right, both of us are mistaken, the notion of being right or mistaken is not applicable to this debate. It may be the case that our conceptions of deep interpretation are so widely divergent that while using the same words, we are talking about different activities. Moreover, even if we exclude this possibility, the reader must decide whether our disagreement can be grounded on a sufficiently large area of agreement. For even if we share sufficiently close conceptions of deep interpretation, our disagreement may be only about the application of our view of interpretation to the particular case we are debating. What from Guyer's understanding of my work reveals an animus against deep interpretation, from my own understanding exhibits a defense of deep interpretation.

Now, if the reader understands our debate as a discussion about the facts of the matter, then only one of us can be right. If it is a discussion about an interpretation of what I have said, implied or suggested, then the fourth possibility deserves to be examined. As a result of his examination, the reader may find that he is confronting two irreconcilable understandings of my work. By opting for the understanding that seems to him most appropriate, he exemplifies the view of interpretation defended in my work.

As a professional interpreter of Kant, Guyer has certainly a clearer view of what Kant said or should have said than Kant's amateur followers. In writing about interpreting, I followed a direction first indicated by Kant, but contributing to Kant scholarship was not my aim. I may have missed the target of Kant interpretation by an inch or a mile: this is irrelevant from the viewpoint of understanding the interpreting activity. We agree on a central point about Kant's views: "For Kant, the claim that if one has in fact made one's own judgment of an object correctly then others who also experience it under optimal conditions can be expected to judge it the same way ... is the content of the judgment of taste, what is meant by calling the object beautiful, not the premise or justification for it." For as Kant wrote: "The judgment of taste ... only ascribes this consensus to everyone [es sinnt nur jedermann diese Einstimmung an]...." I did not quote this passage, but I alluded to it: "According to Kant, judgments of the beautiful are uttered with a universal voice: in saying that a given object is beautiful, I enter the claim that all other reasonable persons would judge as I do; a judgment that agrees with mine is imputed to them." (p. 74)

I must add that I did not assert that universalizability is a premise or a justification for every interpretive statement. Guyer seems to fault me on this issue. After reexamining all twenty-four occurrences of 'premise,' 'justify,' or 'justification' in my book, I could not find what I have contributed to his misunderstanding. Since I did not explicitly deny the view Guyer attributes to me, a translation into his preferred idiom of what I said may have provided grounds for claiming that I implied or suggested what I did not assert.

A full account of my approach to criticism in the arts or a detailed exposition or critique of Kant's views is not within the scope of my book on interpreting. What is in its scope is a discussion of our interpreting activity within certain limits. Interpreting does not have a natural beginning or ending point. If in our examination of interpreting we do not arbitrarily assign a starting point for the interpreting activity, then our notion of interpreting becomes too broad. I have stipulated that interpreting starts when there is a need for an interpretation, when we do not understand what is at issue without further inquiry. There was no need to assign an end point to interpreting. If my stipulation is accepted, then I cannot provide a general account of speech-acts, or the intentions behind the words and deeds of others. Also, my understanding of what is inside and outside of an artwork will be different from what is accepted by others.

According to Guyer, "all that is strictly internal to a poem is a series of black marks on white paper that could be described by a geometer or a series of sounds that could be described by a phonologist." If this is what you consider internal to a poem, then it follows that "Any meaning the object has and thus anything about it that is to be interpreted will in this sense already involve something external to it...." What Guyer calls internal to a poem is from my viewpoint as external to it as the strings of '0's and '1's my computer is producing while I am writing words of a natural language. What is described by the phonologist or geometer is not a poem. To be sure, without what can be described by them, the poem would not exist; without the corresponding machine language, my writing on a computer would not be preserved. This may be of some importance for the ontologist, but from the viewpoint of the interpreter it is irrelevant.

To be sure, phonologists can confirm recurring patterns in the sound structure of a poem. Unless we hear these patterns independently from the phonologist's work, this will not contribute to our interpretation of that poem. It is not enough to recognize the long nasal vowels and soft l-sound in Verlaine's poem. These are ingredients that are easily recognized. Similarly, food tasters can recognize the ingredients of a dish, but this does not imply that they can also evaluate that dish. The evaluation of a poem requires that we bridge the gap between our recognizing its sound structure and our verdict that it is a poem, a good poem or a great poem. In bridging that gap we find what I mentioned before as the key to Verlaine's poem. Once we find that key, we are able to derive the content of that poem from its sound structure. It is precisely this key that poet-translators try to find when translating a poem into another language.

4. Reply to Mary Wiseman

After offering in her own voice an excellent sum of my views on interpreting, Wiseman suggests that my account of "the points of view from which an interpreter can regard what was said or done, her own or that of the speaker or agent or of any rational person ... is curiously incomplete." Since there is no natural end to interpreting, and we can always add to our interpretive commentary, any view of interpreting is incomplete. Still, I was surprised to learn that I am charged with this failing. In at least two contexts I wrote about incompatible viewpoints in interpreting.

The first context is unproblematical, the second raises important questions. Suppose I am confronted with a Nazi who is a firm believer in the eliminationist policies adopted against members of so-called "inferior races." From the viewpoint of all who share his views it is right-and maybe even a duty-to exterminate as many Jews, Roma and Sinti as possible. Now, it is fairly easy to show that not only is he wrong, but also the view that he shares with others is wrong. Their viewpoint is wrong and must be dismissed. For it subdivides humanity into superior and inferior groups, and destines the so-called inferiors to elimination or in a milder form to slavery.

I will introduce the second context by an example. According to the gossip among Philosophy graduate students of the 1950's, Austin once gestured in the direction of a passing colleague on the Berkeley campus, and said to his students "that man is lying through his teeth." The remark was prompted by Austin's failure to convince his colleague that we see physical objects rather than sense-data. What motivated Austin in doubting his colleague's sincerity? The fact that his colleague could not be persuaded by arguments contradicting his views is insufficient for supporting the charge of insincerity. There are incompatible viewpoints that do not admit negotiations. In these cases, adopting the views of the opposing side is not a live option. At issue between Austin and his colleague was an undecidable philosophical problem. Doubt about the sincerity of either side in such a debate requires additional evidence.

Wiseman draws our attention to the fact that if we cannot grasp what was said from the interpreter's viewpoint, we may be able to grasp it from the speaker's viewpoint. She is quite right, and this is precisely what we do when in the large majority of cases we examine another person's self-understanding and come to believe that it is sincere and in accordance with facts. In these cases we are guided by his self-understanding. "Such an interpretation is successful if the interpreter's understanding of the speaker's words agrees with the latter's self-understanding."(p. 46) This will not be of help in the above two contexts, for the interpreter may be unable to grasp what was said from any viewpoint. Let us focus on the first context. I have repeated what was said about inferior races, and this may even persuade you that I have some understanding of what I said - but, is this really the case? Since I am a firm believer in the moral unity of mankind, I find the very idea that there are superior or inferior groups of human beings incoherent. At other times and other places even some philosophers-e.g., Plato and Aristotle-made sense of this idea, but this does not imply that we can make sense of it.[1]

There is an allusion to phenomenalism in the second context. Austin's colleague believed that he saw sense-data, and physical objects were nothing but logical constructs of sense-data, while philosophers sympathetic to Austin's views held that they saw physical objects and sense-data were nothing but philosophers' day-mares. The suggestion that each side must understand the other from the other's viewpoint is not helpful in these cases. But beyond these admittedly rare contexts, is the advice that we try to understand the other from his viewpoint, when I do not succeed from mine useful for solving interpretive problems?

We cannot rely on another person's deficient self-understanding. In these cases the interpretee has only psychological, but not epistemic access to his own words and deeds. "For example, if a child complains 'I have a pain in my hair', we cannot expect to understand him as he understands himself. The self-understanding of the speaker or agent may be insufficient, his vocabulary limited, his judgment clouded. In these cases we set aside the speaker's self-understanding, and we substitute an interpretation of his words that accords with facts known to the interpreter." (p. 47) At one time in our remote past, we have had similar complaints as the child in my example, but this does not help us to understand this child's complaint. By setting aside this child's self-understanding and replacing it by facts about the child and his circumstances, we can provide an interpretation of what he said.

Attributing to the speaker what by our own lights is the speaker's viewpoint, and adopting his viewpoint for our interpretation will not be helpful. (Note that even in this case interpreting starts from the interpreter's viewpoint.) We can always repeat what the speaker has said, but when we paraphrase his sentences, are we sure that we understand him? No doubt, interpreters may be mistaken about some facts that are important for understanding of what was said, they may be insufficiently educated, and they may suffer from various infirmities that prevent their understanding and appreciation of what was said or done. The tone-deaf to music and the unmusical to religion risk ridicule even as amateur interpreters in these fields. Still, in cases when no agreement can be established between a writer and his competent interpreter, we must be prepared for the possibility that the writer's claims are either false or incoherent.

Aestheticians are not professional critics, yet all three contributors to this conversation challenge me to expand on my remarks on criticism in the arts. John Gibson asked for a suggestion about the reading of modernist poetry; I believe I answered his query, provided that it is understood that I spoke with the voice of an amateur.

Paul Guyer challenged the notion of the best available interpretation in the arts from a viewpoint that is incompatible with mine; while I believe he is mistaken, readers who adopt his viewpoint of what is internal and external to an artwork may agree with him, and judge that I am mistaken. Also, he faults me for not explaining why an amateur critic need not search for the best available interpretation. I failed to do so, for I believed that my point will be understood without further explanation. Within our interpretive practices, professionals have a duty to learn about the contributions of their colleagues; amateurs are entitled to plead ignorance of the professionals' work. E.g., professional critics may provide a better interpretation of Racine's or Verlaine's fragments of language than what I have offered. Yet, as an amateur I am entitled to hold on to my interpretation, no matter what the professionals have decided.

Finally, Mary Wiseman suggests in the context of her discussion of a de Kooning painting that "the viewer has to make a case for the sets of intentions the painting seems to her to fulfill. Stern would say that the set for which the best case can be made is the best interpretation of the work. Given different evaluative criteria, different interpretations might count as the best." She got me right. Among the many other topics where she is right, I would like to draw attention to a deeper problem that she raises.

When in my student days in the early 1950's I first formulated my interpretation of the two language fragments, I was as ignorant as Racine or Verlaine about the phonological studies of poetic language. When in the 1990's I became acquainted with some of the scientific literature on the subject-and especially with the work of the Hungarian linguist Iván Fónagy-it was easy to see that my interpretation did not contain anything original. Due to progress in the sciences, what were at one time interpretations became facts. While I did not know it, they were already facts at the time when I first formulated my interpretations.

I must report on another advance in the sciences. I repeated in this paper what I said in 2008 in the symposium preceding this conversation "most poems were written to be heard," but this time I added that "most of us must rely on our biological ears for listening to poetry." At that time I relied on impressions gathered from poets and from the practice of habitual readers of poetry. Now I am relying on recent advances in the neurobiology of reading, and I am inclined to believe that there is sufficient scientific evidence for my claim.

Relying on memory, I claimed that I arrived at my own interpretation of the two language fragments. Do I really know that I wasn't influenced by others who were acquainted with the scientific studies of poetic language? I cannot know that. Hence, it is possible that even at the time when I first formulated my interpretations, I merely engaged in a futile exercise ordinarily compared to reinventing the wheel. Similarly, do I now know that I wasn't influenced by others who knew about recent work in neurobiology, when I confidently wrote that "most poems were written to be heard"? My answer is negative. So, Wiseman is right, "Let the interpreter speak in her own voice and not a borrowed one, and let the object of her interpretation be itself and not a mere projection of her memories and desires, beliefs and doubts, longings and dreams."

5. Envoy

Within the limited space available, I could not reply to all criticisms of my views. Their defense was not my primary aim. In a profession that thrives on disagreement, professional courtesy is expressed by entering our caveat against another philosopher's conclusions, or by reaching his conclusions while rejecting his premises. John Gibson, Paul Guyer, and Mary Wiseman offered what for each of them was the best available interpretation of their respective reading of my book. They have done tremendous work in trying to understand my views and in formulating their critique. For this I am most grateful. They have grounds for demanding the agreement of others, for each of them has satisfied the normative and factual constraints on interpreting. The fact that they do not agree with each other or with me does not diminish their achievements or the importance of their contributions to the debate. For it is precisely such disagreement that makes thinking and writing about interpreting valuable.



Professor of Philosophy (ret.)

Rutgers University

New Brunswick, NJ 08903

Published on July 28, 2010.

    1. For further details, see the remarks about attitudes towards slavery in my book, pp. 84-85 and 168-186.return to text