Validity in Interpretation by E.D. Hirsch

E.D. Hirsch’s book Validity in Interpretation is a classic work in the field of interpreting texts. His book is not specifically related to Koine Greek, nor is it specifically related to biblical interpretation. However, Hirsch’s discussion of authorial intent and genre are important for anyone attempting to interpret any text, including biblical texts.

The summary below was provided by David Stark and is used by permission:


Preface (vii)
· However specific generic classifications are, some ambiguity always exists about “border regions” where genres may overlap.
· This book contributes to general hermeneutics, especially as regards the problem of validity.
· Hermeneutical skepticism questions the right of any humanities discipline to speak of genuine knowledge. All humanities studies are based on interpreting texts, and the validity of initial interpretations is crucial to the validity of larger arguments.
· “Certainty is not the same thing as validity, and knowledge of ambiguity is not necessarily ambiguous knowledge.”
· Two moves are necessary in interpretation: first, the interpreter attempts imaginatively to move into the shoes of an author and attempts to interpret the text (“negative capability”); second, the interpreter checks the interpretation he receives against the author’s actual text to see if the interpretation is valid (“severe discipline”).
· This book’s principles should help interpreters “gain confidence that consensus can be reached by mastering the relevant evidence—whether or not all of it is laid out in print.”

1) In Defense of the Author (1)

a) Banishment of the Author (1)
Recent scholarship has questioned the dictum that texts mean what their authors meant in writing them. In the early days of this critique, it was at least assumed that the texts would mean what their authors’ critics thought they meant if the texts did not mean what their authors meant. Eventually, a theory of semantic autonomy arose, but this theory failed to recognize that “meaning is an affair of consciousness not of words. . . . Whenever meaning is connected to words, a person is making the connection.” When critics propose better meanings for texts than the texts’ authors had, the critics themselves become the originators of meaning. “To banish the original author as the determiner of meaning was to reject the only compelling normative principle that could lend validity to an interpretation.”

b) “The Meaning of a Text Changes—Even for the Author” (6)
It is now widely accepted that texts’ meanings can change. If an author forgets what he meant by something, that does not mean that his meaning at the time was not determinative for what he wrote. Moreover, if an author changes his mind, he must recognize his original meaning in order to make a comparison to assert his changed opinion. In this case, meaning remains the same, but an author’s relationship to that meaning changes. “Meaning is that which is represented by a text; it is what the signs represent. Significance, on the other hand, names a relationship between that meaning and a person, or a conception, or a situation, or indeed anything intangible.” If an author meant one thing when he wrote a text and later decided he meant something else by what he wrote, an interpreter has to decide which meaning should be normative for the text.

c) “It Does Not Matter What an Author Means—Only What His Text Says” (10)
Eliot never complained about people interpreting his writings in ways he did not intend. One may question whether an author’s attempt to convey a meaning may be done imperfectly and the resulting text actually convey a different meaning. Evaluation always tests the match between intention and accomplishment. Public consensus cannot be ultimately determinative for a text’s meaning.

d) “The Author’s Meaning Is Inaccessible” (14)
Of course, we cannot ascertain an author’s meaning with absolute certainty, but his meaning is own encoding system is never completely private. Lack of absolute certainty does not indicate the impossibility of some level of certainty. In any case, the author’s meaning is probably never fully accessible to the interpreter if for no other reason than that the author does not always (if ever) completely know all he is putting into the writing of a text. Moreover, not all meanings can be put into words, which texts can convey. “It is far more likely that an author and an interpreter can entertain identical meanings than that they cannot.”

e) “The Author Often Does Not Know What He Means” (19)
When another person claims to understand an author’s meaning better than the author himself, that person cannot actually live up to his claim. What he may mean, though, is that he understands the subject matter of which the author speaks better than the author understands it. Even in the case of authors who assert that their writings mean nothing in particular, they as authors still determine the meaning of the text. In any case, linguistic signs cannot have meaning for and by themselves.

2) Meaning and Implication (24)

No necessity arises from a text, which compels a reader to make authorial intention an interpretive norm. The interpreter could actualize a different standard if he had a different goal, but only with authorial intention at the center can interpretation be a corporate enterprise. While those who try to discover the author’s meaning in a text may learn something in doing so, those who seek only their own meanings learn nothing new. So, pragmatically, seeking authorial intention is more beneficial than the alternatives. If an interpreter’s claim to having a valid interpretation is to hold, he must be able to subject it to “a genuinely discriminating norm,” of which the only one that exists is authorial intention. All forms of communication and interpretation require that an author’s meaning be determinate and reproducible.

a) Defining Verbal Meaning (27)
“Alice is right to say that Humpty Dumpty cannot successfully make words mean just anything he wants them to.” Sometimes a specific statement has a unique meaning; other times, a particular statement “merely imposes limitations and is not uniquely required for the meaning that is actually willed.” The argument that a mother tongue imposes necessary thought categories on those who have it ignores the great variety of expressions and thoughts actually evidenced by people with any given mother tongue. Although all languages do impose some constraints, they are really limited only by the criterion of sharability. “Verbal meaning is whatever someone has willed to convey by a particular sequence of linguistic signs and which can be conveyed (shared) by means of those signs.”

b) Reproducibility: Psychologistic Objections (31)
To say merely that each person’s interpretation will be different because each person is different confuses mental processes with their objects. Interpreters can fail to understand what authors mean by certain words, but no decisive answer is forthcoming about whether this misunderstanding is necessary. “Meaning is an affair of consciousness,” but limiting one’s observations to the sense impressions one receives instead of speaking about the outside world those observations reflect “is precisely the sort of misplaced sophistication that is found in the psychologistic account of meaning. . . . But the remarkable fact of consciousness is that the objects of its awareness are not the same as the subjective ‘perceptions,’ ‘processes,’ or ‘acts’ which are directed toward those objects.” Perceptions of a thing can vary from one instance to another, but one can still recognize the sameness of the thing perceived (e.g., a table, a phoneme). Because an apparently infinite number of causes may produce the same verbal meaning, authorial intention is apparently still elusive. The inability to distinguish between meaning and significance precludes empirical verification of an interpretation. Yet, in no case, is verbal meaning the same as any rational complex in which it can participate. Finally, those who deny the sharability of meaning cannot (on their own theory) hope to share critique of sharability with others.

c) Reproducibility: Historical Objections (40)
It is quite obvious that we can never recover all the historical data, which would illumine a text from another time and culture. While there are meaningful differences in these dimensions, however, there is also substantial continuity, which makes interpretation possible. Appendix 2 criticizes Gadamer’s defense of radical, hermeneutical historicism. The fact that every generation must interpret texts for itself does not necessarily mean that each generation must understand texts differently from any other. Further, because language and assumptions in any given culture may be highly variable, someone more temporally removed from a text may be in a better position to discern its author’s intention than someone less removed from the text. Of course, we can never be certain that our interpretation is completely correct, and although we are generally more prone to interpret contemporary texts correctly, this maxim does not always hold.

d) Determinacy: Verbal Meaning and Typification (44)
“Reproducibility is a quality of verbal meaning that makes interpretation possible: if meaning were not reproducible, it could not be actualized by someone else and therefore could not be understood or interpreted. Determinacy, on the other hand, is a quality of meaning required in order that there be something to reproduce.” Determinacy sets boundaries for possible meanings for a given statement. Gadamer’s “quantum-leap theory of meaning has no foundation in the nature of linguistic acts nor does it provide any criterion of validity in interpretation.” If meaning can change, an interpreter has no firm basis on which to discriminate between more and less appropriate possible meanings for a given text. Because determinacy typically only defines a range of meaning, the author’s will is required to choose a particular meaning from this range as the actual meaning of the text. Sometimes, context is said to determine meaning, but this statement is slightly incorrect. Instead, context provides a window on other meanings the author has willed, each of which helps to clarify the others. Determination of meaning and interpretation always involves a choice; thus, verbal meaning is not always identical with the meaning of which an author is conscious because authors frequently mean more than they explicitly intend because they conceive of meanings as wholes, which implicitly “mean” their individual parts.

A type has two main characteristics. First, it is bounded. Thus, it resembles a class, but a type may only have one instance, whereas a class typically has more than one instance. Second, a type “can always be represented by more than one instance.” If verbal meaning is sharable, then it is a type.

e) Determinacy: Unconscious and Symptomatic Meanings (51)
That meaning must be bounded and susceptible to validation does not exclude unconscious meaning because “the principle for excluding or accepting unconscious meanings is precisely the same as for conscious ones.” Unconscious meanings are implicitly willed meanings. As such, they are expressed by signs (which is arbitrary and conventional), but they may have additional symptoms unique to a given author, such as mental state and stylistic preferences. Symptomatic meanings are indeterminate and, therefore, should not be confused with verbal meanings. Conscious meanings are like the top of an iceberg, and unconscious meanings are like the part of an iceberg, which remains attached to the top, although it is submerged. Coherency and boundary are merely different ways of speaking of the same principle. Thus, when examining a proposed, unconscious meaning, one needs to examine its coherency as one would do with any other conscious meaning. If a text seems to exhibit one or more, inconsistent, willed types, one should recognize that different, conflicting impulses may be in evidence. “Symptomatic, involuntary meaning is part of a text’s significance, just as its value or its present relevance is. But significance is the proper object of criticism, not of interpretation, whose exclusive object is verbal meaning.”

f) Determinacy: Meaning and Subject Matter (57)
Some terms may be used (e.g., tree) without necessarily “meaning” all that might be rightly included with the term (e.g., roots). There is, therefore, a subjective dimension to how meaning attaches to linguistic signs. Even if someone uses a term without also “meaning” some of its rightful parts, a distinction remains between what the author meant when he wrote and what he may later admit that he should have meant if convinced of his oversight. Thus, subject matter cannot determine implications, the topic of the next

g) Determinacy: Meaning and Implication (61)
No verbal meaning lies on the surface of a text; all meaning must be derived from a text’s signs. Some meanings are dependent on others, and implied meanings are labeled as such merely because they form part of a larger whole. “This array [of the larger whole], along with the principles for generating it, I call the ‘meaning’ of the utterance, and any submeaning belonging to the array I call an ‘implication.’” Determinacy of verbal meaning is, of course, related to determinacy of implications whereby some may be included in verbal meaning and some may be excluded from it. Significance always entails meaning in relationship to something else, and no limits really exist to the number of possible relations in which a given meaning might stand. While types are complete in themselves, they also implicitly include a means of determining whether a given entity (e.g., a square) embodies a given type (e.g., a polygon). If no one were aware of the implications of a type, communicating verbal meaning on this point would not be possible. “Implication belongs to a meaning as a trait belongs to a type. . . . and . . . the generation of implications depends on the interpreter’s previous experience of the shared type.”

3) The Concept of Genre (68)

Speech involves expressing and interpreting meaning. When one focuses on interpretation, one tends to emphasize the linguistic system in which signs for interpretation occur. When one focuses on meaning, one tends to emphasize the determinative power of the author’s will. “[A] word sequence cannot, under the general norms of language, delimit a determinate meaning, and . . . these norms are not sufficiently narrowed merely by reference to a context. Something more is needed, and that additional dimension can be hinted at by referring to the work of Saussure and Wittgenstein.” Saussure emphasized the importance of distinguishing between speech possibilities (langue) and actualities (parole). Wittgenstein drew similar conclusions. Thus, learning to understand utterances is like learning rules for games, but even after the rules are learned, some debate may exist about which game is being played at any given time and, hence, which rules apply. Yet, because we have no definitive rulebook, we must learn the rules by playing the games (i.e., by experiencing and using genres). “Since a type can be represented by more than one instance, it is a bridge between instances, and only such a bridge can unite the particularity of meaning with the sociality of interpretation. . . . [Meaning] must . . . belong to a recognizable type in order to be communicable.” Thus, to communicate effectively, senders and receivers must master general norms for their language and particular generic conventions.

a) Genre and the Idea of the Whole (71)
Interpreters understand utterances with reference to the type of utterance (i.e., genre) they receive. Type expectations are necessary for an interpreter to make sense of the text he receives from an author because he only receives the text linearly rather than all at once.

Thus, even portions of texts must be able to reflect the genre of the whole text to some extent. Understanding of the traits of these types arises from experiencing other instances of the types and observing their individual traits. Of course these individual traits may express more than one genre. Type expectations control interpretation to such an extent that one may well interpret a text wrongly if one assigns it to another genre from the one to which it best belongs. On the other hand, texts generally contain various suggestions about their genre, and only the final (and, hopefully, most accurate) generic hypothesis and interpreter holds necessarily affects his interpretation of the text. At the same time, information like titles and attributions determinative roles in identifying texts’ genres. All these factors make genre identification an instance of the hermeneutical circle.

b) Intrinsic Genres (78)
To unify the text stream he encounters, interpreter has to make some hypothesis about the text’s genre. This generic conception helps guide interpretation and its revision. Speakers must also attend to conventional generic categories if they wish to be understood by their hearers. Indeed, a speaker’s controlling idea of his whole message is itself a generic conception because it implicitly holds places in the utterance open for parts of the utterance of which the speaker is not yet conscious. Additionally, whenever a speaker wishes to communicate, he has to consider his interpreter’s probable understanding of the genre of his communication. Thus, the speaker who wishes to communicate is limited somewhat to using types he shares with the interpreter, because “[u]nderstanding can occur only if the interpreter proceeds under the same system of expectations” as the speaker. Although generic categories may appear to be somewhat ambiguous, preserving a distinction between them and the meanings they convey is necessary because
misunderstanding (of meaning) sometimes arises simply because of generic misidentification. In fact, the sequential nature of communication implies a distinction between genre and because the same meaning can be conveyed by different genres.

Genres themselves help delimit what knowledge is helpful to interpreting their individual instances and what knowledge is superfluous,and one need not necessarily understand the whole before understanding a given part of that whole. After all, changes may occur in one part of the whole, which affect the interpretation of the whole, but do not really affect the interpretation of some of the whole’s individual parts. Thus, intrinsic genre is a “sense of the whole by means of which an interpreter can correctly understand any part in its determinacy.” Context is “a very complex and undifferentiated set of relevant factors, starting with the words that surround [a] crux and expanding to the entire physical, psychological, social, and historical milieu in which [an] utterance occurs. . . . It signifies on the one hand the givens that accompany the text’s meaning and, on the other, the constructions that are part of the text’s meaning.” When faced with difficulty interpreting, we look to other texts only because they are less problematic than the difficulty we presently face. Context provides additional clues to a text’s intrinsic genre. Genre should manifest itself intrinsically in the text, but extrinsic genre may also be imposed incorrectly on a text.

c) Genre Logic and the Problem of Implication (89)
“[W]hen our central concern is validity, we always have to ask whether a particular meaning is or is not implied by an utterance. Intrinsic genre has already been defined as a type, which determines the boundaries of an entire utterance. Thus, meaning is intimately related to intrinsic genre. “We know that a given partial meaning is implied by an utterance, because we know that such a meaning belongs in that type of utterance” (e.g., Socrates, as a man, implies mortality). Implications always depend on the traits of the applicable type, and these traits, in turn, help define that type. Sometimes, “[i]t has . . . been argued that it is artificial to speak of conventions with respect to words and syntax, since within a given language group these elements have ceased to be arbitrary at all. But . . . these verbal difficulties can be resolved precisely because nothing in speaking and interpreting is merely arbitrary, and everything depends on something learned.” That is, however anyone chooses to communicate, the types he chooses to use must be shared.

The mistake of those who would define meaning purely in terms of public norms is not in seeing an extra-personal set of norms as governing linguistic usage but in thinking “that this principle is somehow automatically given to any ‘competent reader.’” Indeed, a correct interpretation may be vague, but a specific interpretation may be incorrect.

Various implications should receive emphasis in interpretation according to how much emphasis their author gave them. “To determine relative emphasis . . . we must have reference to something else that makes the function important, and this something lies at the heart of what genre is [namely,] purpose,” which is here construed as the “final cause” for any utterance. For example, various narrower forms may be classed together as “prayer” or “command” because of their shared purpose(s).

d) The Historicity of Genres (102)
We understand new types of utterances because these types either extend or merge older types. In encountering a new type, one must make an imaginative hypothesis about the conventions for this type until one experiences the type enough to refine this hypothesis. Thus, if an author wishes to invent a new type, he cannot completely break with the types his audience already knows if he still wishes to communicate with them. When instigating this new type, the author may include a number of repetitions simply to provide extra clues to his audience about how they should understand the new type. “The best way to define a genre—if one decides that he wants to—is to describe the common elements in a narrow group of texts which have direct historical relationships.”

e) Variety of Genres and Unity of Principles (111)
Intrinsic genres govern speaking and understanding, and those who wish to understand them must interpret them according to their own specific natures. “To treat a literary text as though it were a document in history or biography is to misrepresent its nature, and such a misrepresentation constitutes a perversion of its meaning. All valid interpretation is thus intrinsic interpretation: whatever one may do with a literary text after it has been understood on its own terms achieves validity only because that preliminary task has been performed. . . . If understanding is always governed by the genre conventions of an utterance, it follows that different types of texts do indeed require different types of interpretation. . . . valid interpretation is always governed by valid inference about genre.

Thus, while the same methods and categories are not universally applicable to all texts, the proper categories are nevertheless always determined by a universal principle—namely, their appropriateness to the intrinsic genre of a text.” For example, all philosophical texts belong to the genre of philosophical texts, but broad textual classifications are often too general to define different types of interpretation. If we define truth as coherence, then contradiction must be absent, but if we define truth as correspondence to a reality, which may include contradictions, then truth may be contradictory. “[T]here is no such thing a s the philosophical interpretation of philosophy or the literary interpretation of literature, but there emphatically is such a thing as the intrinsic interpretation of a text.” Moreover, simply labeling a method as appropriate for the interpretation of a given genre does not make that method appropriate for that genre.

Some genres may permit some (but not all) shifts in vocabulary to have identical meanings. Of course, wordings with potentially synonymous meanings are not necessarily synonymous. Intrinsic types are large, complex types, which govern individual willed types as wholes. “The genre concept turned out to be the principle for determining whether a particular meaning was willed—whether it belonged.”

Some texts like the Constitution or the Bible go beyond the normal limitations of determinedness precisely because of their genre as legal and religious texts. For example, if a law can be made in one time and read as meaning to include a situation, which arises only later, and which the framers of the law could not have foreseen (e.g., legislation about wheeled vehicles translating into legislation about vehicles that perform the role of wheeled vehicles). Because other texts (e.g., Hamlet) do not share these genres, similar interpretive moves are not legitimate, although nothing in the text may explicitly contradict a proposed interpretation.

“Sensus plenior” is one way of speaking about meaning in religious texts that goes beyond what the original, human author could have willed. In this case, the extension of meaning owes its existence to the genre of religious text or to a divine author cooperating with the human author in writing the text. These two authors may have written the same words with different meanings intended for them.

4) Understanding, Interpretation, and Criticism (127)

a) The Babel of Interpretations (127)
Unprincipled, textual commentary has existed from Alexandria to the present day. All interpretations are slightly different, in part, because no interpretation fully describes a text’s meaning. Interpretations may refer to the same meaning, but they may bring out different traits of that meaning. At the same time, tactful interpreters can emphasize whatever traits of a meaning they wish without suggesting a false emphasis. There is, then, something to be said for being aware of several interpretations of a text since this will, generally, help one know the text’s meaning more fully. Some interpretations, however, may be incompatible with one another, but the multiplicity of possibly conflicting interpretations produces difficulty only when one fails to distinguish the kinds of differences the interpretations exhibit. Interpretations, which express the underlying meaning of the text are to be preferred, in whatever language they come, to interpretations whose form may resemble the text but whose meanings may differ from it. “Interpretation” normally refers to explanation and understanding. Here, however, it will refer only to the explanation of meaning, and understanding will be used as a separate category. Encountering different interpretations may have two effects: it may deepen or change one’s interpretation of a text. The first occurs when a new interpretation articulates the same meaning as the one the reader has construed with additional details. The second occurs when a new interpretation is proposed, which presents itself as more plausible than the one the reader formerly held. This principle may be illustrated by two men who view a building from different perspectives. They both view the same building (i.e., meaning is identical), but their different perspectives make them aware of different details (i.e., traits) of the building they see. They can both enrich their understanding of the building by combining their interpretations.

b) Understanding, Interpretation, and History (133)
“[T]he act of speaking implies in itself a projected or imagined act of understanding.” Understanding always requires active construction, as evidenced by the fact that no one can understand an utterance in a language he has not learned. Moreover, one cannot possibly summarize a text unless one has first grasped its contents. “Verbal meaning can be construed only on the basis of its own presuppositions, which are not given from some other realm but must be learned and guessed at.” Thus, “[o]ne cannot understand meaning without guessing or learning the prerequisites to construe meaning. . . . If we do not construe a text in what we rightly or wrongly assume to be its own terms then we do not construe it at all.” While each generation must interpret texts afresh, the meanings of the texts do not change; instead, only the perspectives of their interpreters shift. Yet, no one method or category is the correct one for interpreting all texts.

c) Judgment and Criticism (139)
The account of meaning here is neither narrow nor purist. Significance was earlier defined as the relationship between verbal meaning and something else. Judgment is the ability one has to perceive significance. While the goal of interpretation is understanding an author’s meaning, “we certainly can isolate or at least emphasize a particular goal for our activity.” Essays about texts are “commentary,” and criticism is commentary, which indicates significance.

d) Intrinsic Criticism (144)
Criticism itself is of two types: intrinsic and extrinsic. Intrinsic criticism evaluates a text’s meaning in relation to its own intrinsic genre; intrinsic criticism is, thus, not merely the use of special vocabulary in the act of criticism. Extrinsic criticism evaluates a text’s meaning in relation to anything besides its own intrinsic genre. For instance, if an author wills to write a nonsensical ballad, intrinsic criticism would judge the ballad actually written in relation to the genre “nonsensical ballad,” but extrinsic criticism might judge the ballad according to the “ballad genre.” Thus, extrinsic criticism would not necessarily comport with the author’s intention, but it would be able to describe what the author should have done if he had wished to communicate within the ballad genre. Extrinsic criticism may also be valuable because an intrinsic context may not always be the most helpful one to use when evaluating a certain text for a certain purpose, but one cannot definitely say what extrinsic material will be most helpful for this task until one considers the critic’s goal in performing this type of criticism. Nevertheless, however one evaluates any literary piece, one must always evaluate it as a literary piece if one’s evaluation is to be valid. To evaluate it as, for instance, a mathematical equation, would be to mistake the intrinsic genre appropriate to the text under consideration. Moreover, one must be careful to evaluate the text according to its intrinsic genre (e.g., lyric poem) rather than a more general generic construct (e.g., poem). Genre, however, cannot be used as a straightjacket for individual texts; rather, individual texts have their own individuality and forms of expression. Similarly, a text’s goals and norms are not determined by its genre, but a text is recognized as exhibiting a certain genre partially because of its goals and norms.

Gadamer’s view of interpretation “cannot provide any satisfactory norm of validity,” but one must recognize that this criticism itself is extrinsic because Gadamer was mainly concerned with the historicity of interpretation rather than with validity.

e) Critical Freedom and Interpretive Constraint (155)
“In this section my purpose is to defend the right of literary criticism (or any other criticism based on a broad generic idea) to be as ‘literary’ or ‘unliterary’ as it pleases and still to qualify as objective knowledge and objective valuation.” While critics may have the right to evaluate texts in any contexts they might choose, the standards for their judgments are not so open to arbitrary choice. “But subjectivism is not avoided by following a particular method or adopting a particular vocabulary and set of criteria. The most firmly established method can disguise the purest solipsism, and the more toughminded or ‘objective’ the method appears to be, the more effective will be the disguise.

Objectivity in criticism as elsewhere depends less on the approach or criteria a critic uses than on his awareness of the assumptions and biases that deflect his judgments.” Appropriate, extrinsic criticism is always like intrinsic criticism insofar as this type of extrinsic criticism always considers an author’s purposes in writing. This type of criticism, however, remains extrinsic because the primary reference point for the criticism is the critic’s own set of criteria. An inappropriate, extrinsic evaluation (e.g., reading a ballad as a mathematical equation) can be true and valid like an intrinsic evaluation can be, but such an extrinsic evaluation has little value. “[T]o disagree with purposes the author did not entertain or to praise him for meanings he did not mean is to invite misunderstanding.” In the end, we cannot escape the fact that “almost all judicial and descriptive criticism is predicated on understanding.” The final problem to be considered in this book is validation, which is “the process which shows that in a particular case . . . knowledge [of a text’s meaning] has probably been achieved.”

5) Problems and Principles of Validation (164)

a) The Self-Confirmability of Interpretations (164)
“The activity of interpretation can lay claim to intellectual respectability only if its results can lay claim to validity,” yet certainty of a valid interpretation is fleeting simply because of the nature of the enterprise. Consequently, when an interpreter maintains “certainty” in the face of contrary, possible interpretations, he has probably fallen victim to believing that texts automatically confirm their own interpretations. For example, when Ventris deciphered Linear B, his interpretation was consistent with the data available, but this consistency could merely have arisen because his interpretation came from the data. Only when additional texts, which Ventris had not used to develop his original hypothesis, became available could his interpretation be confirmed. Similarly, patterns and effects of a text observed under one interpretation may appear quite different (though still intelligible) under a different interpretation.

This entrapment in the hermeneutical circle means that an interpreter must always entertain alternative hypotheses, but because not all hypotheses are equally compatible with one another or with the data they claim to represent, the interpreter must also adjudicate which hypotheses are more probably true than others. Too frequently, interpreters are content to avoid this adjudication by either merely integrating different interpretations as best they can or allowing the variety of interpretations to feed their cynicism about finding textual meaning.

This chapter’s purpose is “to describe the fundamental principles that govern the validation of interpretations and lead to objectively grounded discriminations between conflicting interpretations—despite the circularities and complexities which bedevil the interpretive enterprise.” Yet, this chapter will recommend no panacea but will “clarify concepts and encourage a degree of methodological self-consciousness” as it shows how validation standards can be reliable.

b) The Survival of the Fittest (169)
Quotation is the most basic form of validation because it typically functions to show how a given interpretation is not inconsistent with the text being interpreted. By itself, however, quotation, cannot indicate the likelihood of a given interpretation. Interpretation begins and ends with hypotheses about meaning, but right guesses cannot be guaranteed by applying specific interpretive principles. Validation is always relative to alternative hypotheses available at the time of validation. That is, validation does not determine, strictly speaking, the correctness of a given interpretation; rather, it determines that one interpretation is more probably correct than the available alternatives. Thus, “[a]n interpreter is usually deceiving himself if he believes he has anything better to do” than adjudicate the various hypotheses about the meaning of the text he interprets and show how his interpretation is more valid than the other hypotheses. Thus, to be completely validated, a hypothesis must be compared with all available competitors. Moreover, hypotheses stand and fall as wholes; when an interpreter chooses portions from different hypotheses, he essentially proposes another, eclectic hypothesis, which must be validated like any other hypothesis. Yet, “[t]he aim of validation . . . is not necessarily to denominate an individual victor, but rather to reach an objective conclusion about [the] relative probabilities [of the available hypotheses].”

Finally, the distinction proposed between validity and correctness in interpretations does not deny that correct interpretation is possible. It merely recognizes the provisionality of all interpretations, which require re-validation whenever an alternative hypothesis arises.

c) The Logic of Validation: Principles of Probability (173)
Probability is sometimes described quantitatively, but most probability judgments we make in life are made on a qualitative bases like “‘more,’ ‘less,’ ‘very,’ and ‘slightly.’” In this section, we discuss some of the principles of probability judgments as they apply to interpretation.

Implied in every probability judgment is uncertainty about some part of the equation. Probability judgments are hypotheses based on knowns, which relate to unknowns. Thus, a probability judgment may be valid in terms of knowns, but if an unknown becomes known, it may contradict the previous judgment. The probability judgment must incorporate unknowns into a hypothesis based on knowns. This principle implies a second attribute of probability judgments—namely, their reliance on class attributes to describe unknown class members. Ubiquitous characteristics of various objects become defining traits of classes.

Because an unknown is never completely unknown (otherwise it would be unintelligible), one can place it in a class based on its known traits. This procedure does not require complete uniformity among class members; indeed, such uniformity is absent from the subject matter of humanities disciplines. Probability judgments are necessary parts of textual interpretation. This observation is especially evident in the cases of interpretive difficulties when some crux must be understood in light of other parts of the text.

Additionally, identification of intrinsic genre, which bounds meaning, is always a hypothetical exercise. The more narrow the class into which we can put a given instance, the more certain we can be about interpreting that instance correctly. Typical attributes for some child classes, however, may not be typical traits for some parent classes. Consequently, child class traits receive priority for determining what material is most like a given instance. Thus, “[a]nything we can do to narrow the class, such as determining authorship, date, tradition, and so on, will decrease the doubtfulness of our probability judgment—that is, increase its likelihood of being true.”

“Three criteria are decisive in determining the reliability of our guess about an unknown trait—the narrowness of the class, the number of members in it, and the frequency of the trait among those members.” Of course, as a class narrows, its members become fewer, but narrower classes always provide more relevant information than do more class members. Hence, narrower classes are more likely to lead to more correct interpretations, particularly because the act of further defining a class brings one closer to a definition of the unknown member of that class.

d) The Logic of Validation: Interpretive Evidence (180)
Thus, we must distinguish between competing probability judgments based on the evidence supporting each, but accumulating favorable evidence can never prove the correctness of a given hypothesis. Instead, a more reliable method is constructing various tests, which will lead to the rejection of incorrect hypotheses. Unfortunately, with historical material, this type of test is infrequently feasible. In this case, we have to weigh probability judgments against themselves and determine which is the most probable amidst evidence that seems to support various, competing hypotheses. Ideally, this process of weighing probability judgments occurs after one has received all available evidence for all competing hypotheses.

To weigh the competing hypotheses, one must be able to assign a relative weight to the evidence adduced in their favor. The standard for assessing this evidence must be objective because only an objective standard can lead to a publicly compelling hypothesis. First, relevant evidence for this process must affect the relative probabilities of the different hypotheses. Evidence equally commodious to more than one hypothesis is irrelevant for deciding which hypothesis is most likely. Second, relevant evidence will “define the subsuming class [or increase] the number of instances within the subsuming class.” On this basis, we can guess about a trait of an unknown based on the frequency with which known class members exhibit this trait. Thus, the more narrowly one can define the class to which a given instance belongs, the more similar other class members will be to the unknown member and the more relevant their data will be to the interpretation of the unknown member.

Sometimes, however, one must evaluate relative probabilities based on disparate classes. In this context, the criteria of “legitimacy, generic appropriateness, correspondence, and coherence” will help determine probability. Legitimacy and coherence are frequently perfunctory categories since difficulty selecting a correct interpretation from various possibilities arises because more than one possible interpretation explains the phenomena of a text well and is substantially consistent with itself. Still, these criteria are necessary preliminaries. Correspondence seeks to identify the interpretation, which shows the relevance of the most textual traits. Correspondence, however, cannot be calculated by a simple tally because one interpretation might not explain one particular type of phenomenon, which occurs frequently, but another interpretation might not explain several types of phenomena, which occur infrequently. In this case, the first interpretation is more probable because it most reliably accounts for the most types of textual phenomena.

How well an interpretation accounts for a text’s overall class attributes can also help distinguish between competitors. A genre should initially be hypothesized as independently from the actual context as possible by taking into account things like authorship, date, and setting. Moreover, one may consider other evidence external to a class to determine what attributes that class member is most likely to have. These smaller-scale judgments can accumulate to make a larger-scale judgment more or less probable. “When . . . judgments based on . . . two classes are in conflict, we must decide which judgment is the more probable by comparing the copiousness of the subsuming classes and the relative frequency of the predominant trait within the classes.”

Usually the best interpretation of a text will be supported by several probability judgments. When a definitive result is not available, the tendency of the supporting judgments should be consulted. If this procedure does not resolve the difficulty, a clear decision cannot be made on the basis of the information available. “In the course of making any of these probability judgments, the interpreter’s chief concern is to narrow the class” (i.e., to learn as much as he can about the text and external factors related to it).

e) Methods, Canons, Rules, and Principles (198)
Theoretically grounding a discipline should formulate methods, which will produce reliable results if applied properly. Canons of interpretation have existed for ages, and interpretive rules have been multiplied for different types of texts. Yet, no one has ever proposed a practical interpretive convention for all texts. Schleiermacher attempted to delineate several, general interpretive conventions. His first canon holds that an author’s verbal meaning must be intelligible within the author’s own frame of reference. This canon, however, cannot fully account for texts (e.g., laws) whose very purpose extends their meanings to situations outside the author’s sphere of contemplation. Schleiermacher’s second canon was that part of a text must be understood in terms of its surroundings. Yet, because this “canon” describes the way texts are always construed, it is not particularly helpful as an interpretive principle. In the end, Schleiermacher’s real value is not in his interpretive canons but in his detailed qualifications of them. His efforts show that “as a general rule of interpretation . . . there are no interpretive rules which are at once general and practical.” Every practical rule has situations where it does not apply, and attempts at general rules do not provide practical guidance about their implementation.

Yet, practical interpretive canons are valuable as preliminary probability judgments on the way texts normally work. These judgments may, of course, be contradicted by textual particularities, but they tend to hold more often than they do not. Generally speaking, canons are more reliable if they are designed for a narrower application.

Thus, a reliable methodology cannot be built upon interpretive canons. “No possible set of rules or rites of preparation can generate or compel an insight into what an author means. . . . The methodical activity of interpretation commences when we begin to test and criticize our guesses.” Schleiermacher designated these two acts “divinatory” and “comparative,” but he “failed to notice that one function is always prior to the other, that female intuition brings forth the ideas which the comparative male judgment then tests and either accepts or rejects.” Yet, even in the evaluation process, the divinatory element still plays an important role in proposing alternative hypotheses. “The discipline of interpretation is founded, then, not on a methodology of construction but on a logic of validation.”


1) Objective Interpretation (209)

“Criticism” now includes commentary and evaluation, but one can certainly not evaluate a text until one has understood it. Texts are conventional representations, which must be construed before they are of any relevance to the critic. Boeckh made a similar distinction, placing interpretation as understanding a text’s meaning and criticism as construing a text’s meaning in relation to a larger context.

Frege first clearly distinguished between meaning and significance (i.e., the relationship between meaning and something else). As an illustration, he posited a sentence, “A unicorn is in the garden.” He then considered this statement’s verbal meaning in connection with two hypothetical worlds: one where a unicorn was not in the garden (analogous to Frege’s own world) and one where a unicorn was in the garden. In the first context, the statement is true; in the second, it is false. In both instances, however, the statement’s verbal meaning remains unchanged.

a) The Two Horizons of Textual Meaning (212)
If textual meaning changes over time, no objective knowledge of that meaning is possible. Any statement about a text would be valid only for the most fleeting moment and, as such, could not be tested. If a reader’s perspective can determine a text’s meaning, the number of possible meanings equals the number of readers. This situation would be absurd, and fortunately, another way of resolving the difficulties exists.

Eliot located meaning in the total corpus of accumulated literature, but again no textual meaning can be shown to have a relationship with this larger realm of literature until the text’s (that is, the author’s) meaning has first been construed. Two major objections exist to attempts to construe an author’s meaning: (1) the historical determinedness and inaccessibility of this meaning and (2) the private nature of an author’s mental processes. While textual meaning is determined by psychic acts, meaning itself cannot be identified with these acts, as Husserl correctly saw. As an example of the objectivity of meaning, Husserl cited the ability of the same person to perceive an object at different times or from different perspectives, receive different impressions, and still recognize that he was perceiving the same object. In fact, “[a]ll events of consciousness, not simply those involving visual perception and memory, are characterized by the mind’s ability to make modally and temporally different acts of awareness refer to the same object of awareness. An object for the mind remains the same even though what is ‘going on in the mind’ is not the same.” Thus, Husserl distinguished between an intentional act (or expression) and an intentional object (or meaning). In total, Husserl’s analysis indicates that “[v]erbal meaning, being an intentional object, is unchanging. . . . Verbal meaning is the sharable content of the speaker’s intentional object. Since this meaning is both unchanging and impersonal, it may be reproduced by the mental acts of different persons. [Moreover,] any verbal utterance, written or spoken, is historically determined.”

Consequently, Husserl provides some meaningful help in resolving problems of interpretation. In the end, the interpreter’s task is clear: he must distinguish which meanings belong to an author’s verbal intention (or kind of meaning) and which do not. Implications must be determined partly by consulting context, which includes anything that relates to making the text a coherent whole. Yet, implications are not completely defined by what was “in the author’s mind.” To help distinguish between genuine and false implications, one may use Husserl’s idea of the horizon, or a “system of typical expectations and probabilities,” which helps bound a text’s meaning and liberates the interpreter from having to identify precisely what was in the author’s mind. One must strive to identify the horizon of the author’s whole intention because only in this respect can one distinguish which implications are proper to a given willed type and which are not.

An interpreter thus hypothesizes the author’s horizon, for instance, by positing an intrinsic genre for the text. An interpreter must also “familiarize himself with the typical meanings of the author’s mental and experiential world.” In this context, an interpretation may be judged by how well he has identified: (1) the whole work’s horizon and (2) the typicality of his proposed meaning within the whole. Beyond this inner horizon exists also an outer horizon in which a given work can be related to anything else, but critic and interpreter alike must correctly construe the inner horizon to make sense of a text.

b) Determinateness of Textual Meaning (224)
Reconstructing an author’s goals and attitudes can help provide boundaries for a text’s meaning. Texts are not merely free-floating things determined only by linguistic norms, but textual meaning must abide by public norms if it is to be sharable. Moreover, “[an] array of possibilities [of meaning] only begins to become a more selective system of probabilities when instead of confronting merely a word sequence, we also posit a speaker who very likely means something. This section aims to establish that texts represent authors’ verbal meanings rather than being simply ‘piece[s] of language.’”

No single interpretation can exhaust a text’s possible implications. Inclusivism responds by trying to integrate as many interpretations as it can by: (1) viewing one interpretation as a subset of another or (2) fusing interpretations together. The other possible avenue of resolving tension between interpretations (i.e., rejecting all conflicting interpretations) is closed to the inclusivist by definition. The option of fusing interpretations may actually bring together mutually incompatible interpretations. Thus, this approach cannot cogently represent the structure and particular emphases a literary piece is designed to convey.

“Saussure defined langue as the system of linguistic possibilities shared by a speech community at a given point in time.” This system embraces both actual usage and potentially valid usage, which has not yet been actualized. Saussure distinguished this system from parole, or individual utterances. Langue contains the conventions for a language (i.e., its shared norms), and parole is the concrete application of those norms in specific cases. This distinction helps resolve the problem of edited texts whereby one need only determine whether to interpret a text as a parole of one speaker or another, assuming both speakers’ utterances are governed by the same langue. This distinction also clarifies what meaning an utterance has when an author communicates poorly. In this case, it is sometimes argued that the text means not what the author meant but what the linguistic community would understand. Since, however, “only individuals utter paroles, a parole of the speech community is a non-existent [thing]. . . . A text can represent only the parole of a speaker or author, which is another way of saying that meaning requires a meaner.” Of course, determining the parole’s precise meaning is a separate task.

c) Verification (235)
“Since the meaning represented by a text is that of another, the interpreter can never be certain that his reading is correct.” Moreover, any appeal to the “obvious” meaning of the text is susceptible to being simply an interpreter’s own, uncritical reaction to the text.

Thus, the interpreter’s goal must be to show how one meaning is more probable than other possible meanings. He may do this by appealing to legitimacy, correspondence, generic appropriateness, and coherence. Coherence depends on correctly construing a text within its context. If one becomes too convinced about a given, incorrect interpretation, however, he can begin to construe the text ever more wrongly as he becomes ever more convinced of the coherence of his wrong interpretation. Part of a text’s context is the “author’s subjective stance” toward his subject matter, which is not completely knowable, but which may prove helpful in interpretation. This extrinsic data, most properly derived from the text at hand, primarily provides a standard by which an interpreter can check his understanding of an author’s text. This psychological reconstruction must itself also be subjected to all the validation criteria mentioned earlier.

2) Gadamer’s Theory of Interpretation (245)

Gadamer’s Truth and Method has an ironic title because the work serves precisely as a polemic against objective truth and correct method in interpretation. In it, “Gadamer protests that there can be no Methodologie of textual interpretation because interpretation is not, after all, a Wissenschaft whose aim is objective and permanent knowledge.” To assert the contrary is to overlook the historicity of understanding—namely, that no interpreter can move past his own historical distanciation from a text.

Gadamer’s ideas were not new; the terminology he used and explanations he gave are the main value of his work. Gadamer clarified and extended the hermeneutical ideas of Bultmann and Heidegger, and this stream of hermeneutical theory developed into what has come to be called the “New Hermeneutic.” Gadamer grounded his anti-intentionalism primarily on Heidegger’s radical historicism while supplementing it with aesthetics, but he did not appeal to the idea of a collective consciousness. Heidegger’s radical historicism views “historically distorted knowledge as something ‘real’ and ‘phenomenal’ in contrast to academic pseudo-knowledge which is ‘abstract’ and ‘construed.’” Thus, Gadamer thought that one could not truly understand a text’s original conditions. While it has had a profound impact on modern hermeneutics, Gadamer’s theory contains troublesome, internal conflicts and inconsistencies.

a) Tradition and the Indeterminacy of Meaning (247)
Gadamer identified meaning not with authorial intention but with subject matter. He takes this identification partly from Luther, who thought one could not know the sense of a word without knowing its referent. Gadamer, however, failed to distinguish between meaning and subject matter like Luther did. Gadamer argued that written language especially is autonomous and should not be construed as recorded speech. It followed that autonomous written texts had indeterminate meanings, a fact that Gadamer at least partially accepted. Thus, no proposed interpretation could ever properly correspond to a text’s (infinite) meaning, and no interpretation could be judged as being more correct than any other interpretation. On this point, however, Gadamer came dangerously close to eliminating the possibility of meaning since a statement, which can mean anything, practically means nothing in particular.

Gadamer tried to provide some type of norm for evaluating interpretations in the role he gave to tradition. Nevertheless, because the tradition (i.e., history of interpretation) is always changing as new interpretations are rendered, it can provide no stable standard for evaluation.

b) Repetition and the Problem of Norms (251)
Yet, if no reliable standard for validation exists, one questions the merit of even writing a hermeneutics text like Gadamer did. As a way around this difficulty, Gadamer attempted to articulate a way that a text’s individual parts could have repeatable meaning, which could serve as a standard for validation.

c) Explication and the Fusion of Horizons (252)
Because each interpreter is locked in his own historical setting, however, each interpreter always understands a text differently. This conclusion creates some difficulties, which Gadamer attempted to remedy by positing an identity of explication with understanding. In doing so, he fused two acts classically distinguished in hermeneutics and hamstrung his ability to describe how an interpreter understands before he explicates so that the interpreter may subsequently explicate. Gadamer’s solution was to assert that an interpreter’s horizon fuses with the original horizon, and his interpretation is a combination of the two horizons. Nevertheless, this “solution” still implies at least a partial construal of something derived from the original horizon, which radical historical determination supposedly precludes. Additionally, “once it is admitted that the interpreter can adopt a fused perspective different from his own contemporary one, then it is admitted in principle that he can break out of his own perspective,” and this admission lays waste Gadamer’s primary assumption.

d) The Historicity of Understanding (254)
Gadamer’s three main ideas for validation standards (i.e., tradition, quasi-repetition, and the fusion of horizons) are all attempts to bring together past and present while asserting the unconquerable separateness of the two. On the one hand, there is validity to “Gadamer’s insistence that a vital, contemporary understanding of the past is the only understanding worth having and his . . . insisting on the differentness in the cultural givens and shared attitudes between the past age and the present one.” On the other hand, Gadamer’s fundamental mistake is his failure to distinguish between meaning and significance. By distinguishing these two individual processes, although they often conflated in practice, Gadamer could have avoided the self-contradiction in his position. The act of construing meaning is always prior to the act of relating that meaning to something else.

Gadamer himself was content to rest on Heidegger’s argument for radical historical determinedness, but this argument is essentially that modern interpreters cannot truly access past meanings because the past is ontologically unique from the present. One can bypass Gadamer and Heidegger’s radical historicism by recognizing that ontologically different things frequently have some understanding of one another, as do two people having a conversation. Radical historicism may be true, but it is, at its heart, a faith commitment like any other, and it is a faith commitment, which is less plausible than the faith commitment that not all interpretation fail to understand past meanings. On the grounds that somewhere, sometime, ontologically different beings have understood each other, the hypothesis that the same thing can occur for beings differentiated (at least partly) by time is made more likely than its competitor.

e) Prejudice and Pre-Understanding (258)
The most significant point Gadamer makes against the possibility of interpretive objectivity is his assertion that personal prejudice always affects the meaning one construes from the text. This section will argue that “prejudice” actually plays a far more beneficial role than Gadamer assigns it.

Certainly, the whole clarifies the meaning of the parts; thus, to construe the parts, a preliminary conception of the whole is necessary. This first hypothesis must come from the interpreter, since he must have it before he comes to the text. Thus, Gadamer argues, we should embrace these preconceptions in our reading as ways of ensuring textual relevance for ourselves.

This argument depends on equating pre-apprehensions with prejudices when the two categories are not identical. “Prejudice” generally denotes a habitual (and possibly unchangeable) attitude, but a pre-apprehension is merely a beginning hypothesis about the data, which cannot suggest a hypothesis completely by themselves. The data must be construed, and as they are construed, the hypothesis about their relation to the whole may change.

Valid pre-apprehension correctly grasps the author’s meaning, but according to Gadamer, one cannot know before reading a text what the author’s meaning is. Still, the interpreter participates with the author in a system of linguistic conventions, which allow this gap to be bridged. One basic convention is that of genre, a concept that originates with Schleiermacher. The interpreter’s idea of a text’s genre relates the text’s parts to the whole, but this idea ultimately remains a hypothesis like any other idea.

3) An Excursus on Types (265)

a) Self-Identity of Types (265)
This section will discuss several principle concepts for the term “type.” A type is “a mental object or, if one prefers, an idea. The essential feature of a type idea is its ability to subsume more than one experience and therefore to represent more than one experience.” A type (= class) does not subsume instances by abstracting traits from the instances, although doing so would describe the attributes of the class. Rather, consciousness is able to judge the similarity or identity of one thing with another and place that thing in a certain class as when one recognizes an object already observed as being the same thing.

Normally, a type has at least two instances, and these two instances, while distinct, must have some identical attributes for consciousness to subsume them under the same type. Types may be generated either by observation of instance attributes (e.g., pencil) or by simple fiat (e.g., yellow).

b) Verbal Meanings as Types (269)
Locke insisted that types were construed based on previous sense experiences. While types can also arise from fiat, verbal meaning, as a type, is always a learned experience. The inclusion of an instance within a type implies both the identity of the instance’s attributes with the type’s attributes, but it also implies a certain unknown quality to each object, about which one can guess based on the attributes of the type. “As the number of explicit traits [of a type] rises, the area of vagueness diminishes so that fewer different instances can be subsumed by the type.”

Type concepts are indispensable tools for understanding verbal meaning of any particular instance. Like anything else, encounter of a text is always a partial encounter, and experience of that text’s verbal meaning is a type experience precisely because it includes both known qualities and unknown expectations. Thus, types have heuristic and constitutive functions for verbal meaning. Moreover, one can completely understand another person’s verbal meaning, as specific as it is, since verbal meaning is a willed type. On the other hand, the vague areas in that verbal meaning cannot be further specified.

One comment
Leave a comment »

  1. [...] D. Hirsch’s Validity in Interpretation (New Haven, Conn.: Yale, 1967) available on his website, ̔Ελληνιστί, in HTML [...]

Leave Comment