What a Text Means

Ask a congregation what a Bible verse means and nobody will ask you what you mean by mean. Ask a room of interpreters and the question will occupy the afternoon, because the word covers at least four different things. It can point to what the writer was trying to convey. It can point to what the words, as a piece of language, actually say. It can point to what the passage does in the reader who meets it. It can point to what God is doing through the text now. These are not the same, and a discussion that slides between them without noticing produces the argument where both sides are certain and neither is contradicting the other.

The stakes are practical and they arrive fast. If meaning is fixed by the author, then a reading the author would not recognise is wrong, however edifying it is, and a great deal of devotional preaching is in trouble. If meaning is a property of the text as language, the author's biography becomes optional and you may read Paul as you read a poem, which historians will resist. If meaning happens in the reader, then the same passage genuinely means something different to a Guatemalan tenant farmer and a Cambridge don, and the question becomes whether either of them can be wrong.

There is also a specifically Christian version of the question waiting at the end of this lesson, and it will not be dissolved by choosing one of the three answers. If a text has a human author and God as its author, and those two do not have identical knowledge, then whose intention are you recovering when you exegete? That question is old, it is live, and the tradition has never produced a settled answer to it.

What follows gives each of the three positions in the form its own best defenders give it. You should not be able to tell from this lesson which one is being recommended, because the aim is to make you able to argue all three and then choose knowing the cost.

The clearest modern statement belongs to E. D. Hirsch Jr., in 'Validity in Interpretation' (1967) and 'The Aims of Interpretation' (1976). Hirsch's argument is not a piece of nostalgia about authors; it is an argument about the possibility of being wrong. If a text's meaning is not determined by what its author willed, then there is no fact of the matter against which a reading can fail, and criticism becomes the production of new texts under old titles. Hirsch thought that price too high, and he thought most people who claimed to pay it went on behaving as though authors meant things.

The part of his account that travels furthest is a distinction, and a great many arguments turn out to be about it without saying so. Meaning is what the author willed to convey by the words. Significance is that meaning brought into relation with something else: with you, with your church, with the year you are living in, with the rest of the canon. Meaning, on this account, does not change, because what a man meant in the year 55 cannot later become something else. Significance changes constantly, and should. On Hirsch's terms, when a preacher applies Philippians to a congregation in Nairobi, he is not altering Paul's meaning but stating its significance, and confusion between those two operations is the source of most bad interpretation. Hirsch later granted more ground than his 1967 book did. In 'Meaning and Significance Reinterpreted', in Critical Inquiry in 1984, he allowed that application belongs to understanding itself and moved some distance toward Gadamer. How large a retreat that was is disputed: some readers take the distinction to have survived the concession intact, others read the essay as conceding the very point the distinction existed to protect. Read it and judge for yourself.

The theological versions of this position are more sophisticated than the caricature of mind-reading a dead man. Nicholas Wolterstorff, in 'Divine Discourse' (1995), argues for authorial-discourse interpretation rather than authorial-intention interpretation, and the difference matters: the object of inquiry is not the psychology behind the text but what the author counts as having said by these words in this setting, which is a public act, not a private mental state. Kevin Vanhoozer, in 'Is There a Meaning in This Text?' (1998), builds the same case from speech-act theory: a text is something someone did with words, and what was done is recoverable because it was done publicly, in a shared language, for an audience. Walter Kaiser presses the strictest version, holding to a single meaning for a single passage and rejecting the idea that a text carries any sense beyond what its human author asserted.

The strong form of the objection is that this position asks for something it cannot have. We have no access to Paul's mind, only to his sentences; every claim about his intention is in fact an inference from the text and its setting, which means the text was doing the work all along. Defenders answer that the same is true of every conversation you have ever had, and you do not conclude from it that your friends' sentences mean whatever you like.

In 1946 the literary critic W. K. Wimsatt and the philosopher of aesthetics Monroe Beardsley published an essay in the Sewanee Review titled 'The Intentional Fallacy'. Their claim was that the design or intention of the author is neither available nor desirable as a standard for judging a work of literature. Not available, because we do not have the author's mind and often do not have the author. Not desirable, because if we did have a statement of intention, we would still have to check it against the work, which means the work is the authority. A poem, on their account, is detached from its maker at birth and belongs to the public.

This is the position called New Criticism, and its practical discipline was close reading: attend to the words, the structure, the ambiguities, the shape of the thing in front of you, and stop using biography as a substitute for attention. Applied to scripture it produced a genuine gain that even its critics acknowledge. A generation of readers had learned to treat a biblical book as a quarry of sources and a window onto a community behind the text, and to read Mark's Gospel by reconstructing what Mark's church needed. Literary reading turned the window back into a picture and asked what the narrative itself does: how Mark withholds, how the repetitions in Genesis function, why the book of Ruth is shaped the way it is. Robert Alter's 'The Art of Biblical Narrative' (1981) and Meir Sternberg's 'The Poetics of Biblical Narrative' (1985) are the mature fruit of that turn, and they showed that biblical narrators are artists whose techniques can be described.

Paul Ricoeur gave the position a philosophical depth the New Critics had not attempted. His key term is distanciation. When discourse is written rather than spoken, it is cut loose from the speaker's situation, the original audience and the shared world that a conversation can point at; what a written text has instead is what Ricoeur calls semantic autonomy. The text now projects a world in front of itself, and the reader's task is to inhabit that world rather than to crawl back behind the text into the writer's psychology. Ricoeur is often quoted as though this licensed anything, which misreads him badly; his whole programme moves through rigorous explanation and only then to what he called a second naiveté, a recovered capacity to be addressed by a text you have already analysed.

The objection here is that a text without an author is not a text but an inkblot. Words have determinate sense only because they were used by somebody to say something; sever that and you have the appearance of objectivity while quietly installing the interpreter's preferences as the meaning of the words. Defenders reply that language is public property, that its conventions constrain what a sentence can be made to say, and that this is precisely why a competent reader can show another reader to be wrong without knowing anything at all about the author.

The strongest statement of the third answer is a claim about the conditions under which reading happens at all. There has never been a neutral reader, and any account of meaning that assumes one has smuggled a reader in without examining him. Understanding is not content travelling from a page into an empty mind; it is an event in a particular person who arrives already formed, and a theory that cannot describe that person is not describing anything that occurs.

Hans-Georg Gadamer made the case in 'Truth and Method' (1960). Gadamer attacked the romantic hermeneutics of Schleiermacher and Dilthey, with its ambition to understand an author better than he understood himself by reconstructing his mental life. Understanding, Gadamer argued, is not reconstruction but conversation. The reader arrives with a horizon, the whole range of what he can currently see, formed by a tradition he did not choose and mostly cannot inspect. So does the text. Understanding occurs as a fusion of horizons, Horizontverschmelzung, in which neither party is left unchanged. Gadamer also rehabilitated the word prejudice: prejudgements are not obstacles to understanding but the condition of it, since a reader with no prior concepts could not form a question. And he insisted that application is not a step after understanding but part of it, which is why a judge applying a law and a preacher applying a text are doing something interpretation cannot be separated from. Gadamer would resist being filed as a reader-response critic; his claim is about the event between text and reader, not about the reader's sovereignty.

Wolfgang Iser, in 'The Implied Reader' (1974) and 'The Act of Reading' (1978), gave the most careful literary version. A text is not a finished object but a set of instructions with gaps in it. Narratives withhold, imply, leave things indeterminate, and the reader closes those gaps in the act of reading; the work exists in the convergence of text and reader and is not identical with either. This is manifestly true of biblical narrative, which is famously reticent. Why did Cain's offering fail? Genesis does not say. What was Jonah thinking under the plant? The book ends on a question. Hans Robert Jauss added the historical dimension with his horizon of expectations, the set of assumptions a given period brings, which is why the same text can be genuinely illuminating to one century and inert to the next.

Stanley Fish pressed it hardest, and is routinely misquoted. Fish does not hold that a text means whatever an individual wants, because for Fish there are no individuals in the required sense. Readers are formed by interpretive communities whose assumptions determine in advance what will count as a plausible reading, which is why members of a community agree and why disagreement between communities is so intractable. That is not licence; it is a claim about where constraint actually comes from. Roland Barthes, in 'The Death of the Author' (1967), supplied the slogan: the birth of the reader must be at the cost of the death of the Author.

The most serious form of this argument in biblical studies is not literary theory but a historical observation. Readings once held to be the plain sense of scripture, by careful, devout, learned men, were used to defend chattel slavery, the subordination of colonised peoples and the exclusion of women from any public teaching. Those readers were not stupid and they were often better linguists than their critics. What changed was not the Greek. Liberationist, feminist and postcolonial interpreters draw the conclusion that the reader's social location is always operating, and that the only honest procedure is to name it. Elisabeth Schüssler Fiorenza's 'In Memory of Her' (1983) and Phyllis Trible's 'Texts of Terror' (1984) are the standard reference points, and both are rigorous exegetical works rather than manifestos. The objection here is that the argument destroys its own basis: if the reader's location legitimately determines meaning, the text loses the power to correct a reader, and that power is the only thing that made those older readings criticisable. Defenders answer that the correction came through readers rather than around them, since it was people formed by the same scripture, reading from a different place in the world, who overturned those readings, which is exactly the mechanism the position describes; and Gadamer adds that a horizon can only be fused with something genuinely other, so a text incapable of contradicting its reader would not be a partner in the conversation but a mirror, and his whole account requires the text to be able to say no.

Now the Christian version of the problem, which no amount of literary theory settles. The church confesses that scripture is God-breathed and also that Isaiah, Amos, Luke and Paul wrote it, with their own vocabularies, arguments, limitations and dates. B. B. Warfield's word for this was concursus: divine and human authorship running together, without the divine crowding out the human. Grant that, and a question follows immediately. If God knows more than Isaiah did, can the text mean more than Isaiah meant?

The case for yes is not a pious inference, it is an observation about how the New Testament reads the Old. Hosea 11:1 says, 'When Israel was a child, I loved him, and out of Egypt I called My son', and the very next verse says, 'the more I called Israel, the farther they departed from Me'. They sacrificed to the Baals. In Hosea the son is disobedient Israel and the event is the exodus, already past. Matthew 2:15 applies those words to the child Jesus being brought out of Egypt and calls it fulfilment. Whatever Matthew is doing, he is not reporting a prediction that Hosea made. And 1 Peter 1:10-11 describes prophets searching for the time and setting their own words pointed to, while Daniel 12:8 has the prophet say, 'I heard, but I did not understand.' John 11:51 goes further and has a narrator state that Caiaphas prophesied something he did not mean.

The Roman Catholic tradition named the resulting category. Raymond Brown's 1955 dissertation defined sensus plenior as a deeper meaning, intended by God but not clearly intended by the human author, which is seen to exist in the words of scripture when they are studied in the light of further revelation. Protestants have proposed close relatives: typology, in which an earlier person or event is patterned so as to anticipate a later one; and figural reading, revived from Erich Auerbach by Richard B. Hays, in which the meaning of an earlier text becomes visible only in retrospect. Hays, in 'Echoes of Scripture in the Letters of Paul' (1989), argues that Paul's way of reading is not an embarrassment to be quarantined but a hermeneutic the church should learn.

The case for no is not obscurantism either. Its defenders argue that once a text may mean what its author did not say, no reading can be excluded, and every allegorist in history has appealed to exactly this principle. Walter Kaiser holds that a passage has one meaning, that the human authors knew more than modern scholars credit them with, and that appeals to a fuller sense are usually a way of avoiding the harder work of showing the connection from the text. G. K. Beale takes a different route to a conservative result, arguing that when you examine the apostles' citations in their full Old Testament contexts, including Hosea 11 as a whole with its promise of return in verses 10 and 11, they turn out to be respecting context rather than overriding it. Richard Longenecker, in 'Biblical Exegesis in the Apostolic Period' (1975), concluded almost the opposite: that we may share the apostles' convictions about Christ without being able to reproduce their exegetical methods, because they read from a revelatory standpoint we do not occupy. The volume 'Three Views on the New Testament Use of the Old Testament' (2008) sets Kaiser, Darrell Bock and Peter Enns against each other on precisely this, and it is the fastest way to see how far apart serious evangelical scholars remain.

Do not leave this lesson with a resolution, because there is not one to have. What you should leave with is the ability to say where a disagreement actually lies. When two people argue about whether Isaiah 7:14 is about Hezekiah's day or about Christ, they are usually not disagreeing about Hebrew. They are disagreeing about whether a text can carry a meaning its human author did not hold, and until that is said out loud the argument cannot move. Naming the real question is most of the work. The rest of this module is about the tools for the part that comes after.