The Difference Between an Original and a Copy

The requirement to produce an original has a reputation out of proportion to its actual reach. It applies where the contents of a writing are the thing being proved, and not to every occasion on which a document appears in a case. Most of the time a duplicate does the same work, and the exceptions are worth knowing precisely because they are narrow.

When the original is actually required

When contents are in issue. Proving what a contract said, what a letter contained, what a photograph shows. The writing is the evidence rather than a record of it.

Not when the writing merely happens to exist. A witness can say a payment was made without producing the receipt, because the fact is the payment and not the wording of the receipt.

Not for collateral matters. Where the writing concerns something peripheral, the requirement gives way rather than derailing the hearing over a minor point.

Recordings and images count as writings. The rule reaches recordings and photographs, which surprises people who think of it as a paper rule.

Summaries are a separate route. Voluminous records can be proved by a summary, chart or calculation, provided the underlying material is made available to the other side and, if the decision-maker asks, produced. That is a practical accommodation rather than an exception, and it depends on the underlying records themselves being admissible.

What counts as a duplicate

A counterpart produced by the same impression. Carbon copies, and anything struck from the same matrix, which is the old form of the idea.

Any accurate mechanical reproduction. Photocopies, scans and photographs of the document, which is what the category means in practice now.

A printout of stored data. Where information is kept electronically, the printed or displayed output is treated as the original rather than as a copy of it.

Not a retyped version. A transcription is somebody's reading of the document and carries whatever they misread, whatever they silently corrected and whatever they left out. It is evidence of what the transcriber understood, which is occasionally useful and never the same thing as the document.

Not a translation. A translated document is a new writing and is proved with a certification from whoever translated it.

When a duplicate is refused

A genuine question about authenticity. Not a formal objection, but a real reason to doubt that the original said what the copy says.

Circumstances that make it unfair. Where a partial copy is offered and the rest of the document would change its sense.

Where the copy is illegible in the material part. A reproduction that has lost the figure or the signature is not a reproduction of the point in issue.

Where the original is available and simply not produced. The absence of any explanation is itself the problem, and an explanation for a missing original is straightforward to give.

Where the process of copying is itself in question. Selective scanning, cropped images and edited exports each raise a question that the copy by its nature cannot answer, because the missing part is exactly what would settle it. The remedy is the original, or an explanation of how the copy was produced and by whom.

Public records and official copies

Public records are proved by certified copy. The original stays in the register, which is where it belongs, and the certified copy is the standard route.

Certification is by the custodian. The officer who holds the record attests that the copy is what the record contains, and signs in that capacity. A certification by anybody else, however senior, does not perform the same function, because the point of the exercise is that the person certifying is the person with the register in front of them.

Foreign public records have their own route. Authentication is by apostille or by consular legalization, covered under documents issued abroad.

A certified copy is not evidence of truth. It establishes the contents of the register and nothing whatever about whether the register is right. Where the entry itself is doubted — a misspelled name, a reported date nobody checked — the certification adds nothing to the argument and the fight moves elsewhere.

Where no certified copy can be obtained. Testimony from somebody who examined the record becomes admissible, on proof that a copy could not be got by reasonable diligence.

The practical effect of the modern rule is that duplicates go in almost always, and the argument moves to weight. That is the right place for it: a legible photocopy of a document nobody disputes is as good as the original for every purpose the hearing has.

Where the original does matter, it matters early. A question about authenticity raised on the day is far harder to answer than one raised in correspondence months before, when the original might still have been produced.

For records held electronically, the concept of an original has quietly dissolved. What exists is stored data, and what is produced is one rendering of it. The useful questions are about the export rather than about which copy is the first one.

Anybody holding a document that might matter should keep it, and should keep it in the form it arrived in. The cost of doing so is nothing, and the cost of not doing so is an argument that could have been avoided entirely.

The rule is also worth invoking sparingly. An objection that the original has not been produced, made in a case where nobody doubts the copy, reads as obstruction and tends to be treated that way. Raised where there is a real question about what the original said, the same objection is one of the strongest available, because it asks for the one thing that would settle the point. Changes made to a document after it was signed raise a further question, taken up in alterations.