Signatures are obsolete: Replacing the illegible squiggly line.
Adventures in UX, Law, Forensics, and the Meaning of Things.
Abstract: In this article, Accordium co-founder Barrington Russell discusses the origin and legal basis of hand-written signatures; their failings in the modern world, and how his company iterated through a number of technological alternatives to improve upon pen and paper solutions.
Does your handwritten signature actually mean anything?
When it comes to certifying your legal intentions, the illegible squiggly line (or more politely; your signature) is a relic that has endured for more than a millennium.
Seemingly impervious to the progress of time and technology, why does the venerable thrash of pen on paper seem so resistant to change — and why should we do something about it?
It could be imagined that there is something magical about a signature; as if our deftly-drawn scrawls of ink are imbued with a higher legal power to at once identify and bind their author. But do these arbitrary marks actually have any intrinsic legal value? Looking closer, the answer isn’t quite clear to the everyday signatory.
Examining the origin of this outmoded ornament begins with a question:
What is a legally-binding signature anyway?
When making an agreement recognized by law between two parties (a contract, in common parlance), the law generally considers the function of a signature to have three parts:
1. To provide evidence of the signatory’s identity;
2. To show evidence of their intention to make a signature, and;
3. That the signatory accepts the contents of the contract.
Signing paperwork is such a fundamental legal act, that almost every document of commercial importance is signed. Despite this, the artifact of the signature itself is rarely afforded much attention.
To find out why, we have to understand the difference between the Form and Function of a signature.
Form vs Function
In a technical sense, the ‘reference implementation’ for signatures has, since about the Sixth Century, been that of the signatory’s name, written by their own hand, on parchment. This is known as a ‘manuscript signature’ and is the same illegible squiggly line practiced by schoolchildren the world over.
Many analogs and alternatives have been invented in the intervening time, such as initials, crosses, phrases, wax seals, and rubber stamps. But in all cases, courts have been able to resolve whether these marks constitute a valid signature by drawing analogies to the manuscript signature.
The reason this works is that when making a contract, the law is principally concerned with the function of the signature: what it means and what it does. This function is so universally understood by both lawyers and non-lawyers alike, that in practice it is never called into question.
Accordingly (and perhaps surprisingly) there is almost never a specific requirement for the form of the signature, so long as its function is represented and understood.
“In other words, what it is and how it looks are entirely irrelevant, so long as it represents the functions of identity, intent, and acceptance.”
That’s worth reiterating, because it is fundamental to our understanding of why the form — the shape we draw — has no legal meaning on its own.
The truth is that your illegible squiggly line is imbued with no legal properties arising from the specifics of its shape or form. The function however, is everything.
In other words, what it is and how it looks are entirely irrelevant, so long as it represents the functions of identity, intent, and acceptance.
With so many ways to represent a signature, and with such an abundance of modern devices ready to copy and distribute media, it would seem prudent to harden these forms against misuse.
What changed? Why is this a problem now?
Compared to the Sixth Century when signatures first arose, in the 21st Century we are all ultra-connected super-merchants and hyper-lawyers; entering into potentially hundreds of contracts a year, as we sign for credit card receipts, take out subscriptions, rent vehicles, receive packages, make deals, and sign documents in the office and online.
There are also legal and economic incentives today that promote a culture of litigiousness that did not exist in the past — in some countries, famously so. (We’re looking at you, ‘Murica).
Primarily though, the problem stems from the internet, and the way in which documents are rarely based on paper any more.
Where historically we placed tacit legal faith in applying the form of our illegible squiggly line to a parchment, today we are rarely afforded this opportunity. The modern reliance on digital representations of what was once real, and the ensuing divorce from that physicality, leaves a question:
How best to transform the function of the manuscript signature into a useful digital form?
In other words, since nothing about writing or drawing on a screen is unique (binary information can be copied), what meaningful thing can we do on a computer that replaces signing a piece of paper?
Since form is irrelevant and the law doesn’t differentiate between ink and pasting JPEGs onto PDFs, we clearly need a better method to document the evidence of identity, intent and acceptance — otherwise anyone could claim that a stolen scan of my signature, acquired improperly, is a valid legal instrument.
Therefore we argue that the problem with illegible squiggly lines in 2018 — where so much of our personal and business landscape is defined by digital documents,multiple online identities, geographic separation, and frequent and litigious deal-making— is that they have become outmoded by the very functions they aim to satisfy.
Further, they are just plan inconvenient. Since most of our transactions occur on the screen, we would need a way to somehow get the paper into the screen. That’s more cumbersome than it needs to be, usability-wise.
Systems built on paper signatures are easily defrauded and highly perishable records of history.
When tasked with comparing scribbles on a tiny receipt, a barista frothing morning lattes becomes a poor guardian of your legal identity— but that’s the innocent end of the problem.
Multi-million dollar legal battles have been lost on the basis of signature fraud and inadequate validation. Remember that copying a signature onto a piece of a paper by any form you like is a perfectly acceptable method of supplying the necessary legal function.
I think that’s a lot faith and dependency to put onto an illegible squiggly line, don’t you?
What is the fundamental nature of the problem?
Principally, it is that the form of the squiggle is dumb. It has no intrinsic ability to collect data to prove its own validity (a branch of forensics) other than to distinguish its own shape, which is in itself meaningless.
Removed from the context of its parent paper, and cast adrift in digital space, our reliable Sixth Century companion starts to lose applicability and meaning.
Drawing dumb curly shapes is therefore a fundamentally insecure way to certify your legal intent on a contract.
And then there’s that convenience aspect again. Until recently (and probably sometimes even still) we’ve been downloading PDFs, printing them out, signing them with a pen, scanning them back in, then emailing them back!
How obtuse that in 2018 we’re often still forced to jump through this redundant digital-analog-digital hoop.
Technology shows a better way
Computers on the other hand are very good at automating the collection of forensics — the scientific data required to support cases in court, and they’re also very convenient to use.
Remember that when signing a contract, the law recognizes the function of the signature as establishing identity, intent, and acceptance of its contents.
These three dimensions are the legal meta-data attached to a signature.
What we need, then, is a method to create a form on a computer that is smart enough to collect its own meta-data. A ‘smart-form’, if you like.
At Accordium, part of our software is used to sign documents electronically — and when designing our signature component, we wanted to account for each of those three meta-data dimensions.
As I explain how we address each of them, I hope the reader understands the methods by which these ‘smart forms’, or ‘smart signatures’, are measurably superior than their antique manuscript cousins.
Identity
When building a better signature, the first dimension we sought to smarten was identity. How can a signature prove its own authorship?
Our initial indicator of authorship is via someone’s email inbox. We dispatch invitations to sign contracts to known email addresses, and only the owner of that email address can author the signature.
As a first line of defense, knowing that the signature for Barrington Russell originated from a request sent to Barrington Russell’s email, and clicked from a link inside Barrington Russell’s inbox — we can make a reasonable assumption that the ownership is good, so long as the email address is not compromised.
Secondarily, we use a software tool called the Accordium Forensic Engine to capture an array of meta-data about the user session that originates from the email invitation; such as IP address, geolocation, and type of device. The engine can also record some actions taken on the screen, which assist in capturing intent. I discuss this in the next chapter.
Finally, during the signing process we compute hashes for both the contract text and the signature form with its associated meta data, then use asymmetric public key cryptography to uniquely link the document to the signer. This action simultaneously creates a tamper-proof ‘seal’ on document, through which any post-facto changes to the document will yield different hash values, invalidating the contract.
This sounds complicated — and it is — but through the above system, we create a method of identifying the owner of the signature form, and linking that signature to the contract. In this way, the form of the signature (its shape on the page and resultant computed hash) has an integral part of proving its own identity.
So far so good, and already an improvement on the illegibly squiggly line!
Acceptance of Contents
Accepting the contents of a contract is the second dimension of the signature function that we want to improve. It is also the simplest.
The illegible squiggly line already makes some statement about the author’s acceptance by virtue of its presence on the physical document. That’s certainly a powerful indicator, but we can use technology to improve the odds further.
In addition to placing the signature form on the page, the Accordium Forensic Engine can infer whether the document was read in its entirety by measuring things like scroll behaviour, time spent on each page, and cursor movements — hallmarks of a human scanning a page. This can be summed together into an evidence payload.
When the smart signature is embedded in the page, it is associated in our system with the evidence payload from the user who created it.
Secondarily, we explicitly require the user to accept the contents by checking a box and recording their choice.
Finally, since the contents of the page are hashed into a tamper-evident seal, the use of the smart signature carries an innate ability to validate whether its parent document was the same one that was accepted by the author when they signed it.
Also an improvement over the illegible squiggly line.
Intent
The final dimension where computers help us modernize is with capturing intent. In other words, did the person who signed the contract mean to do so in an attempt to become legally bound?
I discuss intent last, because it’s the most interesting, and also a concept where Accordium uniquely differentiates itself.
Of course intent is a phenomenon that is internal to the mind of the author. Without a magical brain-scanning machine, we cannot be certain of someones intentions. Nevertheless, we can be reasonably sure that action conveys intent — and so recording what someone did is arguably a suitable indicator of their intention to do so.
When building our ‘smart signature’ then, we should seek a method to record what someone does while creating it. In doing so, and by introducing enough friction elements to the signing process, we can be reasonably sure that it was performed deliberately, and within the rules specific to our environment.
Friction elements are hindrances inserted into a software user journey, to deliberately arrest a user’s progress. One example might be an ‘Are you sure?’ button. By requiring a specific affirmative action, they act as mediators of intent — guaranteeing that the user is aware of what they’re about to do, and that they mean to do so.
In our case, we record the response and user behaviour on three occasions: once before the signing process, and twice more at the end. Through this we can demonstrate that the user accepted the premise of signing online, and that their electronic signature constitutes an agreement to be legally bound.
One final interaction completes the signing process; operating the TurnKey — something unique to Accordium — and something I’ll discuss in detail at the end.
So asking the user to explicitly acknowledge that they understand what they’re doing, as recording their behaviour as they acknowledge it, is central to providing evidence of intent.
Building a Better Signature
Now that I’ve discussed how our architecture can improve upon the legal function of the illegible squiggly line by applying technology, we can set about designing a new type of on-document form — the appearance and shape of the signature. Remember, it can be anything.
For our purposes, we wanted to create something that satisfied the following criteria:
1. Convenient and enjoyable to use frequently;
2. Identifiable as analogous to a written manuscript signature;
3. Legible at all sizes;
4. Modern enough to highlight the outdated-ness of the illegible squiggly line.
We progressed through a number of versions on our way to the current solution. Some didn’t work! Let’s take a look:
Exhibit A: The Starting Point — Illegible Squiggly Line
Pencil on paper, coffee-stained, and dumb as heck. Our point of origin, extensively discussed, leaves a lot to be desired. No background processes, no analog-to-digital converter, no security and little support for forensics.
Exhibit B: Scrawl n’ Bawl
The first attempt, and the method still preferred almost exclusively by our competitors, is this travesty. It’s a very literal interpretation of the problem. Draw you signature on the screen.
It’s what we call ‘the scrawl n’ bawl’.
Why? Because you’re presented with a tiny canvas on which to draw (or rather, attempt to draw) your signature with a mouse or trackpad. Frustrated, you try again. And again. And again.
Eventually resigning yourself to a spider scrawl, you cry about how useless you feel, and move on — just that little bit more miserable for the experience.
The usability of this system is, quite frankly, shit.
This method is neither enjoyable nor convenient to use, and barely represents a good attempt at your real signature. It’s also totally illegible and does nothing to manifest the auspices of the future.
Since the UX of drawing on a screen is so poor, we reasoned that eliminating it completely was the logical option — so remembering that a manuscript squiggle is not required by law to represent a signature, we did just that.
Exhibit C: The Quill
The Quill was an early attempt at modernizing the appearance of a non-drawn text signature. In this version, the user could type their name into the block and have it appear on the contract.
At the time we felt it was modern, usable, and a cheeky middle-finger to everything that represented the illegible squiggly line. It was the legal answer to punk rock. But proud upstart that we were, our customers soon straightened us out!
We still use the quill in our company logo now, but its heavy-handed application to the page, unfamiliar usage, and overly futuristic font drew a lot of criticism from our users.
Although we knew this mark was legal, no measure of futurist explanations could reassure our customers, who were used to drawing curly shapes, that it was legitimate.
We were forced to backtrack, because our users didn’t believe it (yet).
For readers who’ve got this far, you will by this point understand that there is nothing special about the illegible squiggly line.
But few day-to-day users understand this, and so the mystique of pen on paper survives. The common assumption endures that if some curly shapes aren’t made, or if you don’t have the option to careen your finger over the trackpad like a drunken ice-skater, then it’s not real. The mind is stubborn.
And so we had to accept that it was a little early, and that the time wasn’t quite right for the modern Quill and its friends. A hybrid compromise was required.
Exhibit D: Introducing the TurnKey & Stamp
The solution we settled upon has two parts. One is a novel dial that users turn to generate the signature form. The second is the application of that form onto the page — the ‘stamp’.
The TurnKey is another friction element device, which is a fancy way of saying that it stops the user and forces them to understand the outcome of their action — before applying some type of affirmative action to complete the process.
By turning the dial, a more traditional handwritten signature of their name is revealed, and they are prompted to use this as their signature. The software asks them to acknowledge that this mark will represent their legal identity, and then the entire process is captured by the Accordium Forensic Engine before being embedded in the document evidence package.
Critically, it’s not a process that can be copied outside of Accordium, and neither can it be done by accident or by someone else. In this way, our TurnKey has served the function of evidencing identity, intent and acceptance.
Importantly, there’s no shitty drawing required. It’s fast, convenient and enjoyable to use on a wide range of devices, and offers a degree of tactility that feels like you’ve done something real.
Finally, the generated form — the ‘stamp’ — is applied to the page, and then the document is hashed, encrypted, and sealed.
We feel our hybrid stamp is both modern and fresh, but also representative of the old way of signing. It has the right balance of tradition and futurism, and contains a number of identifying elements.
One element is the beginning of the unique hash that is computed from the TurnKey procedure, second a small security padlock and branding line, and finally an array of coloured discs.
The discs serve as a type of watermark, but more importantly indicate on the contract where some interaction has taken place.
On a piece of paper it’s easy to see the difference between the original printed text, and a signature drawn on it with a pen — the colour, ink sheen, style, and appearance are different. But on the screen it can be hard to differentiate between the original PDF, and something added to it. So we add these subtle indicators to point out where supplemental marks have been added to the original document.
The feedback from our customers showed a much higher preference for this type of form.
Conclusion
I have discussed at length the legal aspects of what a signature is, the relationship between its form and function, and how we can use technology to improve the odds when guaranteeing identity, intent and acceptance.
Drawing curly shapes on a page feels natural and comes with a sense of gravitas. But this action isn’t well suited to the screen, and so superior methods will be developed to replace it.
Some degree of user education is required though. Getting people to let go of their entrenched idea of what a signature is, or must be, in order to be real may be a struggle. I believe that rather than an all-or-nothing approach, a stepping-stone hybrid route like the above will get us there, one step at a time.
The electronic signature industry is young, but regulatory frameworks have already been created in Europe (EIDAS), the United States (E-Sign Act), and in many other countries around the world. Provided certain technical considerations are met, these regulations provide a legal basis for the types of implementation detailed above.
It is our hope that in time, businesses and individuals alike will come to a greater understanding about the nature of signatures, and how technology can replace the old illegible squiggly line.
Accordium is a machine to speed up sales departments. By sending contracts to the right people at the right time automatically, you can generate revenue faster and with less effort. Barrington Russell is co-founder and Chief Product Officer.
