How I Read: The Rules Behind the Exegesis Papers

Every verse-by-verse paper on this site used to open with a few pages explaining how I read. The rules didn’t change from paper to paper, so the explanation didn’t either, and a reader who had seen it once had to scroll past it again. This page is that explanation, written once. The papers now start closer to verse one and point here.

If you only want the short version: two rules govern everything, a grading scale keeps me from overstating what I’ve found, and five further commitments fall out of those two rules once you hold them consistently. What follows is each of those, with the test cases that made me trust them.

Rule one: Scripture interprets Scripture

No word in a passage gets to mean what a modern dictionary assumes it means. It gets to mean what the rest of the Bible shows it means by how the word is actually used.

That sounds obvious until you see how often it’s skipped. The standard Hebrew and Greek lexicons were largely written by men who already believed Genesis was about the origin of matter and Revelation was about the end of the planet. Their definitions carry that belief in as if it were data. A definition is a guess about usage. It has to be checked against usage, and when the two disagree, usage wins.

So every lexical claim in these papers was tested the same way: take the Strong’s number, sweep it across the whole canon, and report what the word does everywhere it appears, with counts. When a paper says a word “occurs four times,” that count was run against the text, not copied from a reference work. When the sweep went against me, the paper says so.

Two examples from Genesis 1 show why this rule isn’t optional.

“Create” (bara, H1254) is covenant language. Isaiah uses this exact verb for God’s making of Israel: “he that created you, O Jacob… he that formed you, O Israel” (Isa 43:1); “I have created him for my glory” (43:7); “the creator of Israel, your King” (43:15). And in the one place outside Genesis where the Old Testament narrates a bara of “heavens and earth,” it explains itself in the next line: “Behold, I create new heavens and a new earth… for, behold, I create Jerusalem a rejoicing, and her people a joy” (Isa 65:17–18). The same verb, three words apart. To create new heavens and earth is to create a city and a people. If that’s what Isaiah’s bara of heavens and earth means, I have no right to insist that Genesis’s bara of heavens and earth must mean quartz and hydrogen.

“Divide” (hivdil, H914) is the priestly verb. It structures all of Genesis 1: light from darkness, waters from waters, day from night. It’s also the verb God uses for the act that makes Israel his: “I have separated you from the peoples, that ye should be mine” (Lev 20:26; see also Lev 10:10, dividing holy from profane). Genesis 1’s world is built by the verb that builds a holy people.

Neither of those was cherry-picked. They’re the load-bearing verbs of the chapter, and Scripture’s own usage points them at covenant, people, and city.

Rule two: audience relevance

A text means what it meant to the people it was written to. Genesis wasn’t written to explain geology to the nineteenth century. It was written to a covenant people to tell them who their God is, who they are, what went wrong, and what he swore to do about it. Its questions are their questions: not “how did matter arise” but “how did God come to dwell with a people, on what terms, and why are we outside his presence.”

The same rule holds at the other end of the Bible. When a first-century letter tells a first-century church that certain things are “shortly to come to pass,” it means what it says to them. The burden is on anyone who wants to suspend that, and the burden is heavy. The imminence language appears in every New Testament author and every genre. The writers expected it, and said so. The only live question is whether they were right.

State the rule precisely, because a loose version invites an easy jab. The rule is not “fulfillment within the hearers’ lifetimes.” Zechariah’s hearers waited five centuries. The rule is that the fulfillment must land inside the story those hearers were living in. Haggai promised the men rebuilding the second temple that the nations would come to that house and fill it with glory (Hag 2:7, 9). Any fulfillment dated after that house ceased to exist in AD 70 isn’t late. It’s impossible. The covenant reading asks for the same patience Simeon had: waiting inside the same story for the consolation promised to it (Luke 2:25–32).

A reading that has a prophet interrupt a comfort oracle to deliver an astronomy bulletin has failed this rule before it ever reaches rule one. The exiles weren’t worried about the durability of the sky. They were worried about whether they were still a people.

The grading scale

Rules one and two tell me how to test a reading. The grading scale tells you how much weight I’m putting on the result. Every verse in these papers gets a hypothesis, a test, and one of four grades:

• CERTAIN. The text or the canon states it outright. Denying it means denying an explicit statement.
• PROBABLE. Canonical usage strongly favors it. Alternatives exist but are strained, and I say where the strain is.
• PLAUSIBLE. Open. Usage permits the reading but doesn’t push toward it. I hold it, but it carries no weight.
• IMPOSSIBLE. Ruled out by grammar or usage.

Two rules govern the grades. First, the thesis of a paper never rests on anything below PROBABLE. Second, no verse gets conceded while a PLAUSIBLE covenant reading still stands. That second rule doesn’t award me the verse; a PLAUSIBLE verdict counts for nothing in the argument. It only refuses the opposite error, which is treating a verse that merely leans as a verse that’s lost. A futurist is welcome to the mirror rule. It costs me nothing, because none of these cases is a chain that snaps at its weakest link. Each is a set of independent lines that converge on one place, and it stands if a few of its PROBABLE pillars stand.

The grades run both directions. The opposing reading gets graded on the same scale, at full strength, the way its best defender would state it. And where a hypothesis of mine failed testing, the paper reports the failure and drops the hypothesis. There are discarded hypotheses in every one of these papers. That’s what makes the surviving grades worth something.

On the site, the grading apparatus mostly stays in the background so the papers read as prose. But it’s how they were built, and when a paper says a reading is probable rather than certain, that’s a real distinction and not a hedge.

What follows from the two rules

Hold those two rules without exception and five further commitments come with them. I list them because in practice most public exchanges collapse into a method dispute within three replies, and it helps to be able to point at the method directly.

The apostles’ citations are evidence about the Old Testament, not a devotional overlay on it. When a New Testament writer says what an Old Testament text is about, that identification is data. Paul calls Genesis 2’s man and wife a mystery about Christ and the church (Eph 5:32). He reads Psalm 19’s heavens as the preaching of the gospel (Rom 10:18). If we don’t listen to Paul, we’ll miss what’s going on in Genesis. This is only rule one applied across the Testaments, and preterists already grant it against futurism every time they read Matthew 24 by Daniel and Joel instead of by the newspaper. The wall between “Old” and “New” is artificial. Paul preached “none other things than those which the prophets and Moses did say should come” (Acts 26:22). And the apostles didn’t invent the habit. Jeremiah reads Genesis 1 (Jer 4:23). Isaiah re-issues “create heavens and earth” as the making of Jerusalem (Isa 65:17–18). Ezekiel reworks Eden (Ezek 28; 47). Hosea and Job cite Adam (Hos 6:7; Job 31:33). The covenantal reading of Genesis is the canon’s own reading of itself, running unbroken from Moses through the prophets to the apostles. It isn’t a preterist overlay on either end.

Creation and curse must match in kind. Whatever kind of thing the creation of Genesis 1–2 is, the curse of Genesis 3 has to be the same kind of thing, because the curse falls on exactly what God made. You can’t read the creation as material and the curse as covenantal without a reason in the text, and there isn’t one. The same discipline applies inside any passage: a prophet doesn’t switch from covenant to cosmology mid-sentence unless the grammar makes him.

Non-concordism. The text isn’t trying to line up with a scientific description of the world’s beginning or its end. Every young-earth, day-age, and gap reading of Genesis tries to make it line up. So does every futurist reading of Revelation. Both are doing the same thing at opposite ends of the book. Concordism is a real, named category in the science-and-faith literature, and the covenant reading is the one account that declines it at both ends for the same reason: the text was never about that.

Pattern first, then map. Establish what a word or image means from its full distribution before you carry it forward to a disputed passage. Never the other way around. Most bad readings of Zechariah 14 or 1 Corinthians 15 start by deciding what the passage is about and then finding the vocabulary to fit. The order matters.

Fulfillment can look nothing like the expectation and still be fulfillment. Jesus said so about Elijah. Malachi promised Elijah would come before the day of the LORD. A man in camel’s hair came, and Jesus said that was Elijah (Matt 11:14; 17:12), though he looked nothing like what anyone expected and most people missed him. “It didn’t look like what we expected” isn’t evidence against fulfillment. Jesus already ruled on that case.

What this looks like in a paper

A paper opens with the question it’s answering and who it’s written for. Then it names two or three facts about the passage that are checkable in an afternoon and that break the opposing reading on their own, before verse one. Then it walks the verses. Then it assembles the reading, answers the strongest opposing case on its own terms, states where the argument stops, and gives the bottom line.

That “where it stops” section is not a formality. A case that hides its soft joints doesn’t deserve to win, and I’d rather you find the weak places in my own account than in someone else’s rebuttal.

The papers built this way so far: Genesis 1–3, with a shorter version; 1 Corinthians 15; Zechariah 14; and Isaiah 51. If you read only one, read Genesis. Everything else on this site is downstream of it.