How can I tell if an autograph is real?

An adult has signed their own name tens of thousands of times. It has stopped being an act of drawing and become an act of muscle memory, performed in roughly one to two seconds without watching the pen. Everything useful about a genuine signature follows from that speed.

A fast line tapers. It starts thin because the pen was already moving when it touched down, thickens through the downstrokes where the hand pressed, and thins again on the way out — often finishing in a whip-thin tail that leaves the paper in mid-air. Curves flow into one another without flattening, because the hand never had to stop and re-aim.

Somebody copying a signature is doing something completely different. They are reproducing a picture, and that happens at drawing speed — five to fifteen times slower. Slow ink lays down a line of almost even width. There is no pressure story. Curves flatten at the points where the copyist paused to check the model. The pen is placed on the paper and then moved, which leaves a blunt, rounded, ink-heavy start instead of a flying one, and the same thing at the finish.

This is the single most useful thing you can learn to see, and it costs nothing but a magnifier and attention. Check 04 in the guide covers it with a side-by-side animation, and check 05 covers the terminals.

Before any of the stroke analysis matters, establish that a pen was involved. Printed, stamped, and digitally reproduced signatures are extremely common, and in a photograph — even a good one — they can look entirely handwritten.

Under a ten-times loupe the difference is usually obvious. Printed signatures break up into a regular pattern of dots, or show the soft halo an inkjet leaves at the edge of a line. They sit in the same plane as the rest of the image, often underneath the surface gloss, and they never indent the surface. Real ink sits on top, soaks into the fibres of paper, and has an edge that varies along its length.

Raking light helps here more than magnification does. Put a single lamp low and off to one side, so the light skims across the surface rather than hitting it flat. A pen leaves a groove. A printer does not.

This one check ends more items than any other, and it takes seconds. The writing instrument, the ink, the paper, and the photographic stock all have to be plausible for the claimed signing date.

Ballpoint pens did not reach American consumers until October 1945, and were not in common use for some years after that. Felt-tip markers arrived in the 1960s; the Sharpie fine point launched in 1964. Glossy resin-coated photographic paper is largely a product of the 1970s and later. A 1930s studio portrait signed in Sharpie is finished as a question before you examine a single letter.

The reverse warning counts too. Ink on a genuinely old surface should have aged along with it — sunk into the fibres, faded in step with the paper around it. Fresh, glossy, sharply black ink on a brittle eighty-year-old page is its own kind of mismatch. Check 09 includes a chronology of writing instruments you can hold an item against.

Signatures change across a life. Fame compresses them, repetition abbreviates them, and age, injury, or illness degrade them. A genuine 1978 signature should look like 1978 — not like the careful full version the same person produced in 1954.

So gather three to five signatures you can actually trust, dated within a few years of your item. Illustrated auction archives, museum and library holdings, and facsimiles in reference books are all usable. Then compare construction rather than outline: which loop closes first, which direction a capital is entered, whether the t-bar goes on before or after the next letter. These habits are the part of a signature its owner cannot easily change and a forger rarely notices.

Two mistakes are worth naming. The first is judging a late signature against an early example and rejecting something real — this happens constantly. The second is accepting a forgery copied from the single most-published example of a signature, which is exactly the example forgers work from and the one specialists recognise on sight.

People sign for legibility and convenience — on a light area of a photograph, clear of a face, at a size that suits the surface. Forgeries turn up in places that flatter the fake instead: dead centre across a dark background, oddly small on a large print, or positioned to avoid the difficult part of the image.

Provenance is next. Useful provenance names people, places, and dates that somebody else can confirm: a signing event with a witness, a ticket stub, a dated letter, an estate inventory, a dealer who watched the pen touch the paper. Stories are not provenance, however charming they are. If a history exists only in the seller’s description, treat it as a description.

A certificate comes last, and it is worth much less than most buyers assume. It records an opinion, it can be printed by anyone, and certificates from reputable houses get counterfeited and attached to fakes. Read how to tell whether a COA is legitimate before you let one reassure you.

Working through all of this is worth doing, and it will stop you buying a great many bad items. It will not tell you that an autograph is genuine, and it is important to be straight about why.

Three failures are close to invisible at a kitchen table. A secretarial signature was written by a real person at real speed, so it passes every physics-based test. An autopen puts real ink on real paper with a real pen. And a genuinely skilled forger produces writing that behaves correctly under magnification. Each of these is identified by comparison against a large library of known examples, by someone who works with that library daily.

So treat the result as a triage decision rather than a verdict. Keep three lists — what passed, what failed, and what you could not judge — and take all three to a professional if the item’s value justifies it. A clear failure early on saves you a fee. A clean pass is the point at which paying for an opinion starts to make sense. Work through all twelve checks, then decide where to have it examined.