Attachments, Exhibits and What Travels With a Filing

Every attachment is a small promise that the reader will be able to find the right page. The promise is broken by the same handful of failures every time, and each of them is avoidable at the cost of a few minutes at the point the filing is assembled rather than a few hours when somebody asks which document was meant.

What an exhibit needs to be an exhibit

An identifier on the document itself. Marked on the exhibit, not only in the index. A page carrying no mark is a page that cannot be placed once it has been separated from the bundle, and pages do get separated.

A reference in the supporting document. The witness statement or application refers to the exhibit by its identifier at the point where it matters, so that a reader can move between the two without searching.

A description in an index. Enough to identify the document without opening it: what it is, who made it, and what it is being offered to establish, in one line each.

A physical or electronic association. Tabs, a binding, a single combined file. Loose pages in a folder are not attached to anything and will not stay in order through two readings.

Consistency of pagination. One sequence running through the whole bundle, so that a reference to a page number is unambiguous even where exhibits carry their own internal numbering.

Numbering that survives handling

One sequence per filing. Restarting the numbering for each witness, or for each category of document, produces three documents all called exhibit four and a correspondence about which one was meant.

Never reuse a number. If an exhibit is withdrawn, the number goes with it. Reusing it means that everything written about exhibit nine before the change now refers to something else entirely.

Numbers rather than descriptions. References in the body should use the identifier, because a reference to "the bank letter" fails as soon as a second bank letter enters the bundle.

Sub-numbering for related documents. A run of statements from one account is easier to handle as a single exhibit with internal numbering than as fifteen separate exhibits.

Say what the numbering means. A one-line note at the front explaining the convention costs nothing and prevents a reader inventing a different interpretation of it.

What goes wrong with attachments

The exhibit was never marked. So the page is genuine, relevant and unusable, because nobody can establish that it is the page the statement referred to.

Two versions of the same document. One clean, one annotated, both attached at different points, and the difference becomes the subject rather than the document.

A partial document. The first two pages of a five-page letter, attached because those were the pages on the desk, which raises the question of what the other three said.

An attachment nobody can read. A photograph reduced to a thumbnail, a scan at an angle, or a spreadsheet printed so small the figures are illegible, which is covered in why screenshots prove so little.

An untranslated document. Attached in its original language with no certification, which in most proceedings means the reader is entitled to disregard it entirely.

Electronic bundles, and what changes

Pagination becomes essential rather than helpful. In a single combined file the page number is the only reliable coordinate, because internal numbering scattered through the exhibits contradicts itself.

Bookmarks replace tabs. A navigable index inside the file does the work that projecting tabs did, and its absence in a bundle of several hundred pages makes the bundle close to unusable.

Text has to be searchable. A bundle assembled from photographs of pages carries no text layer, so nothing in it can be found except by scrolling through the whole thing.

File names carry meaning. Where exhibits are filed separately rather than combined, the file name is the identifier, and a folder of files called scan001 defeats the exercise entirely.

Metadata travels with the file. Which is usually helpful and occasionally not, and is worth understanding before sending, as what metadata actually records sets out.

Assembling a bundle that works

Start from the list, not the pile. Write the index first, from the argument, and then find the documents. Working from the pile produces a bundle organized by what happened to be in the drawer.

Include what is unhelpful. A bundle containing only supportive documents invites the question of what was left out, and the answer is usually more damaging than the document would have been. The same reasoning runs through explaining a gap honestly.

Check every reference. Each mention of an exhibit in the body should be followed to the exhibit itself, once, before the filing goes anywhere. This catches most numbering errors in a single pass.

Keep an identical copy. The copy retained has to be the copy sent, page for page, or every later discussion about what the other side received becomes unresolvable.

Record how it was sent. The transmission record, the covering letter and the date are part of the filing, and become the evidence in any dispute about whether it arrived.

Attachments are treated as clerical work and are not. The decision about what goes into a bundle is a decision about what the decision-maker will read, and it is made by whoever assembles it, usually without much thought about the consequence.

The most damaging failure is also the quietest: a bundle that contains only the helpful documents. Nobody objects at the time, and the omission surfaces later in a form that is much harder to answer than the document itself would have been.

For anything likely to be read by more than one person, or to be read more than once, the effort spent on identification pays for itself immediately. A reader who can find a page is a reader who is still thinking about the argument.

The electronic version of all this is the same exercise with different tools. Bookmarks instead of tabs, a text layer instead of legibility, file names instead of markings, and the same underlying question: can somebody else find the page that was meant. Where the document is a form rather than a free-standing instrument, its own instructions govern what belongs with it, per reading a Form Against Its Instructions.