This is how I read, how I test what I read,
and the rules I will not break to win an argument.

Things work when they run the way they were built to run.

I spent a working life as an electrician. A circuit does not care what you believe about it. Put it on the wrong leg, load it past what the conductor was sized for, land a neutral where a ground belongs — it will not argue with you. It will get warm, and then it will fail, and the failure will look like bad luck to everyone who did not watch it being built.

Nothing in that is unique to wire. It is the plainest thing I know, and it holds wherever I have tested it: a thing operated outside its design does not break immediately. It breaks eventually, and it breaks somewhere else.

That is the whole of it. Everything here is an attempt to apply that one observation honestly — to language, to text, to the way people read, and to my own conclusions first.

Seven rules I work by.

These are not opinions about what a text means. They are constraints on how I am allowed to get there. If I break one of these to reach a conclusion, the conclusion does not count — including when I like it.

1. The text before the tradition.

   What a passage has come to mean is evidence about the people who read it. It is not evidence about the passage. I read what is on the page first, and I say plainly when the two have parted company.

2. A word is not a concept.

   Words have ranges, not lumps. Pouring every possible sense of a word into one occurrence is the commonest error in popular teaching — it has a name, illegitimate totality transfer, and once you have seen it you cannot unsee it. Each sentence gets to mean what its own sentence needs.

3. No word is the “equivalent” of a word in another language.

   A translator chose it, in a particular century, for particular reasons. That is a decision with a history, and the history is part of the evidence. I name the choice rather than smuggling it in as an identity.

4. Mark the noun and mark the verb.

   They are not the same word wearing different clothes. Collapsing them is how a text that describes something people did gets quietly rewritten into something people held.

5. Exegesis is not verbal demonstration.

   Drawing the meaning out of a passage on its own terms is one act. Setting passages side by side to exhibit a continuity is a different act, with a different warrant. Both are legitimate. Calling the second one the first is not, and most of the teaching I spend my time correcting does exactly that.

6. A study sheet is not a source.

   Someone’s collection of quotations is a finding aid. It tells you where to look. It does not tell you what is there, it is not citable, and I do not stand on anything I have not read in its own place.

7. Follow a drift before you fight it.

   Words move. Pointing at where a word has been is useful; insisting that present speakers secretly still mean it is not. I show the distance travelled and let the reader weigh it. Force applied to a language produces an equal and opposite reaction — and the people who try to legislate usage lose, and are resented for trying.

What I will not do.

The line on the mark is not decoration. Know what you do not is the load-bearing part of the method — the discipline of holding the edge of your knowledge in view while you work near it.

• I will not build a case on a word’s origin alone. Where something came from is not the same as what it now carries.
• I will not claim more than the text will hold, and I will say so at the point where the evidence stops rather than at the end.
• I will not cite a source I have not read in its own place.
• I will not argue against a position I cannot first state fairly in its own defenders’ words.
• I will not present my conclusion as the text’s plain sense when it is my reading of the text.
• I will not treat an unresolved question as settled because settling it would be convenient.

Where the writing lives.