On (digital) Signatures, Verification, and Authenticity

Where to start?

Several years back, I wrote my master’s thesis on the changing nature of writing and authorship given the new realities of online publishing platforms (blogs, etc.). If memory serves, it was an overly optimistic piece (or at least a lot less pessimistic than I am now), much in the vein of the majority hypertext and digital writing scholarship of the late 90s – early 00s. One of the topics that I touched on, through my readings of Derrida, was that of the signature.

To wit:

By definition, a written signature implies the actual or empirical non presence of the signer. But, it will be claimed, the signature also marks and retains his having-been present in a past now or present [maintenant] which will remain a future now or present [maintenant]. thus in a general maintenant, in the transcendental form of presentness [maintenance]. (Derrida, Limited Inc 20).

Digital signatures, however, while attempting to carry the same connotations (and legal implications) as the written signature are both more and less. The legal frameworks for digital signatures have developed over the last few decades such that I can now initial or type my name and the date on a form as a form of a legally binding signature. I can also affix a scanned image of my written signature to digitally produced or digitally transferred documents. Further, touch screen and stylus technology now allow for digital written signatures. Each of these have their own implications. The power of the signature was meant to be the implied (deferred) presence of the signatory based on the actual having-been present necessary to perform a written signature. This is no longer necessarily the case. With the first example – anyone could type my initials into a document and thus sign in my stead. And while this remains a crime in the way that forging a signature has always been, at no point is my having-been present required. With the second, my having-been present is required to create the initial scanned signature but afterwards the image can be affixed by myself or anyone with access to the image. The third example – the digital written signature – is the closest to the written signature and seems to require much the same having-been present state of physically signing. Perhaps technology has brought us full circle.

This brings us to the related topic of verification. Online verification (used on services such as Twitter, Facebook, reddit, and other online fora where some (if not all) must prove themselves to be who they claim to be) presents itself much like a signature. “The blue verified badge on Twitter lets people know that an account of public interest is authentic.” And to receive such a badge, several things are required. Yet there are some interesting and notable differences between verification (especially regarding fora that are primarily text-based). While the requirements of verification would seem to authenticate the individual or individuals responsible for an account (their legal existence), they do not authenticate specific posts.

One example of this would be the many cases of social media managers “going rogue” and publishing content that is deemed contrary to brand messaging. Whether the posts were pre-approved or not being irrelevant in the face of backlash. Another example would be the accounts of celebrities or other people of “public interest” that are primarily written by unnamed (non-signatory) representatives or publicists except when specifically initialed by the verified signatory themselves. A further example would be those bots or automated accounts that claim to be “real” people despite the obvious demonstration of algorithmic behavior (who is the true signatory in these cases? the bot or the programmer?)

A final compelling example is that of Donald Trump. Trump is officially verified under both his own name and that of the office of the President. However, it is generally believed that he does not write many of the tweets that appear under his image and verified “signature.” The account of the office of the President is believed to be written primarily by a staffer and there have been several notable analyses of his primary account that indicate that it is not the work of a single author. Indeed, there is a twitter bot (@thetrumpwatcher) that analyzes each of those tweets to determine both their mood and if they are, in fact, authored by Trump himself (the speculation being that many are not).

This presents a complicated situation. Can we consider this a situation of a surrogate speaking in the President’s name as is a common and longstanding practice? Not exactly. When a Press Secretary stands at a podium and speaks for the President, we are never under the impression that the Press Secretary is the President. Indeed, this is instead an extension of the New Yorker comic sequence about dogs online. internet_dogtumblr_njzlcpxjy91qav5oho1_500

The account is verified, thus is speaks in the name of the President (with attendant “authenticity”) and yet it cannot be understood to be operating in the same context as a surrogate (of which there are many other equally verified Twitter accounts). Moreover, the fact that the verified and authenticated and thus “authentic” voice of the President is often not speaking the actual words of the President does not help to dispel rumors that there are powers behind the throne or fears that the account (accessed via an unsecured Android device) might be hacked. After all, how could we tell?

What seems to be one result of the attempts to verify and authenticate “accounts of public interest” is a further decline in trust and in authenticity itself. Who decided that corporations are in the public interest? Who is determining the interest? Of what public? And if we can only tell that the President is actually writing because of the persistence of typographical and grammatical errors (because a too polished presentation is obviously false, obvious a con), we then lose the shortcuts and conveniences that allowed us to trust public speech at all. Everything becomes branding, PR, propaganda, or advertising. This wouldn’t necessarily be a problem if the public could then be counted upon to verify and authenticate information for themselves (through the myriad of new means that digital/networked/online tools and environments offer). However, that does not yet seem to be the case. Instead of opening up language to its inevitable polysemy and differance, to the inability to communicate without noise and interference (even in person, even with the face to face) we are getting increased ideological fragmentation, tribalism, and partisanship. Instead of celebrating quality news and diverse voices, we are falling for more advertising and propaganda. Instead of critical thinking, we are stuck with covfefe.