Issue 11 · Thaw

The Problem With Starting Fresh

A fresh start has no dependencies because they have not been written down yet.

The new notebook has no crossed-out measurements. The new folder has no duplicate filenames. The new project board contains only deliberate columns. A new plan can appear simpler than an old one for the same reason an empty workbench appears efficient: none of the work has reached it.

Sometimes the reset is justified. A system may be badly designed. A material may have failed. A project may belong to a goal that no longer exists. Keeping every prior choice in the name of continuity produces its own kind of waste.

The problem is that starting fresh discards evidence along with clutter.

The household project records for 321 Steele Street are full of sequences that are inconvenient precisely because they are real. Plumbing work needs to precede a ceiling restoration if both occupy the same cavity. Water management links gutters, rain barrels, grading, drain tile, foundation conditions, and an alley outlet. A new sink is not only a sink; it is supply, drain, location, wall access, and the future work that may need the same route.

An empty plan could make each item look independent. The building would correct it.

Software and archives make the same correction less visibly. A file has links, history, assumptions, and names used by other files. Rebuilding the structure may improve it, but moving information into a clean hierarchy can also sever provenance. A sentence that once recorded a decision becomes an orphaned fact. A failed approach disappears, making it available to fail again.

Freshness is therefore a poor measure of clarity.

A marked-up drawing may be clearer than a clean drawing because it records which dimensions were checked. A reused board may be more legible than new lumber because its holes reveal the former load. An old project note may be awkward and still contain the only explanation for why an attractive option was rejected.

Continuity does not require preserving the whole mess in active view. It requires making deliberate decisions about inheritance.

What remains authoritative? What has been superseded? Which constraint still exists? Which part is history rather than instruction? Where did a number come from? What must happen first?

These questions produce a different kind of cleanup. Instead of pretending the work has no past, they reduce the amount of past that must be carried forward while preserving the parts that still govern the next action.

This can feel slower than opening a new document. It is slower for the first hour. It may be much faster by the third week, when the new system begins rediscovering old dependencies.

The attraction of the reset is not only organizational. Possibility is easier to admire before it has been priced, scheduled, contradicted, or asked to fit through an existing doorway. A new system returns the work to that interval. Everything can still belong. Nothing has yet become the part that must wait.

That relief is real. It is also temporary.

The alternative is not endless patching. A sound revision can be substantial: rename the structure, archive obsolete material, consolidate duplicates, replace a failed component, redraw the plan. The difference is that revision keeps a chain of custody. It can state what changed and why.

In a building, that chain may be visible as capped pipe, patched masonry, an old opening, or a junction box placed for future access. In an archive, it may be a decision note and a date. In either case, the prior work remains available to explain the present arrangement.

A fresh start promises freedom from the old problem. Continuity asks a less pleasant and more productive question: which parts of the old problem are facts?

Those facts will enter the new work eventually. It is cheaper to invite them at the beginning.

Sources