Flogging In The Marketplace Of Ideas
Will Continue Until Morale Improves
An endless outpouring of misinformation characterizes 21st century media; how can its irruptions of irrationality be curtailed? Or is “free speech” paramount; do we just have to accept that (as every generation throughout history has discovered) it’s going to hell on a sled?
A Wretched Figleaf
Defenders of “free speech” — particularly those stolid in advocating for the merits of freely expressed bigotry, defenders of the virtue of parading vile or ill-founded opinions — often approvingly cite the idea that the best response to speech is “more speech”.
This is a shorthand reference to the “Counterspeech Doctrine”, as stated by US Supreme Court Justice Louis Brandeis in 1927. The core of it is that:
If there be time to expose through discussion the falsehood and fallacies, to avert the evil by the processes of education, the remedy to be applied is more speech, not enforced silence.
In other words, the doctrine assumes that misconceived or awful ideas can be effectively repudiated; and after all, over time the needle has shifted radically on what “vile” opinions consist of. People who hold|weaponize terrible, bigoted ideas use this assumption as a fig leaf, fully aware that there are various modes of failure that counterspeech is subject to.
Settlement Out Of Court
The key phrase in the doctrine is “if there be time”. There rarely is. Speech is not, usually, delivered according to the idealized parameters of a court: opposing positions counterbalanced, arguments allotted adequate time to run their course, conclusions reached through rational analysis and debate.
In the primary theatre of debate, the media, there are various distorting forces which mean there isn’t the capacity for “falsehood or fallacy” to be exposed merely through argument.
First Force: Gatekeeping
Firstly, access to media, and particularly traditional media with its broad reach, is generally aggregated to establishment privilege. Money talks, on the whole — or at least has editorial oversight. Bias is innate, because participation is not egalitarian, and some voices are never heard. Consider the highly skewed coverage of “trans issues” that The Times delivers:
Relentlessly negative coverage, admitting no counterspeech: this is nakedly propagandistic, fuel to the fires of a moral panic. (And reflective of broader systematic biases in media coverage).
Second Force: Grappling
Secondly, audience attention is a much more limited resource than content.
As media bandwidth is driven to effective infinity, retaining an audience requires attention-grabbing delivery. The result seems to be a monotonic increase in content extremity: the need for controversy increasingly usurps a regard for validity, mere argument substitutes for investigation. Terms of debate are chosen for their ability to generate heat, not light. Sensationalism wins, becoming the default mode of presentation.
Third Force: Gaming
Thirdly, media does not have robust reputational systems or oversight mechanisms: the result is endemic hornswoggling.
Alongside traditional “spin”, the message amplification techniques used online — astroturfing, bot posting, microtargetting, click farming, etc.— are used to create an echoic environment, a manufactury of consensus. The products of social media, personalized mosaics of contextless reshares that supposedly depict our interests, are shaded with tiles of propaganda.
Red Card
All of these might be summarized as aspects of a single problem: debate being subject to partisan refereeing. Outside of an idealized court, the mechanisms intended to scaffold impartiality are feeble or overwhelmed or corruptible.
Counterspeech cannot provide an effective riposte where it is drowned out by such distortions.
Actions Don’t Speak Louder Than Words
There is another fundamental issue with the notion of refutation always being achievable through counterspeech.
The counterspeech doctrine presupposes a view of speech as propositional in nature: it is rooted in the idea that speech works to outline a “position” — something amenable to discussion, something that can be bolstered or rebutted through the application of arguments. It is a debating society’s view of language, a philosophical traditionalist’s view of language as a medium for logical hair-splitting.
But not all speech is propositional. There is a very different view of language we can take, one first expounded by the philosopher J L Austin in the 1950s. Austin described language as being comprised of speech acts.
Speech acts are all the things we do with words, the actions it lets us carry out. Examples: salutation, swearing, christening, interrogating, pleading, commanding, requesting, promising, chanting, praying, joking, announcing, advertising.
These kinds of actions are clearly not propositions. And many of them are not amenable to being incorporated in debate. If someone Sieg Heils in your face, there’s no discussion to be had. It’s an act of salutation demonstrating allegiance to fascist ideals, a signal of hostility to non-fascists: counterspeech can’t provide any counterbalance. There’s no refutable statement there; no more could one refute the arm motion used.
(In several countries, including Germany and Austria, performing this act is illegal, and can result in jail time).
The fact that counterspeech only works as a coherent response to certain kinds of speech acts is recognised in the penalization of hate speech in many countries (the US doesn’t recognise the general category of hate speech, but does have similar, although much more limited notions of incitement and fighting words). And there are other limits on free speech relating to fraud, defamation, and obscenity, even in the most liberal regime.
But the argument for making the scope of “enforced silence” as narrow as possible recognises the risk of introducing partisan refereeing to speech generically — because any legal mechanism that allows speech to be suppressed would immediately work to suppress the speech of those battling for civil liberties, those agitating for change to the status quo (who do not, of course, work solely through polite “propositional” argument). Enforced silence is indeed a tool of authoritarianism, and needs to be guarded against.
A War Of Words
The protection of “more speech” is not, however, universal: there is an extensive domain of speech, one we encounter every day, where “more speech” is directly harmful. Extensive legal limitations on it (“enforced silence”) have been applied to general approbation over a period of centuries.
That domain is advertising, aka “commercial speech”. Limits on what can be expressed through advertising and how it can be expressed do work to everyone’s clear benefit; and this is not generally contentious.
Maybe this is perceived as different because adverts obviously use specific speech acts that are radically different from those of debate. Advertising consists of an endless series of reiterated assertions and contra-assertions of efficacy or pleasurability, each brand trying to shout its rivals down: in short, sloganeering, to generate the kind of ambient familiarity called “brand awareness”. Slogans are assertions, not propositions: they are not amenable to discussion, not open to rebuttal through logical debate.
(I am here distinguishing raw assertions from propositions; the latter I take as being expected to function within a flow of argumentation in which they can be adjudged true or false, while assertions do not exist within such a context).
Allowing the untramelled expression of advertisers clearly doesn’t lead to an unfolding of truth; it just creates a jumble of words.
The absurdist embodiment of counterspeech in advertising is hinted at in Charles Dickens’ 1851 article that discusses bill sticking, the profession of pasting up posters. The pasted bill was of course one of the key channels for advertising before the advent of electrical media.
Here is Dickens’ description of a derelict building, a site favoured by bill stickers:
Below the rusty cellar-grating, crumpled remnants of old bills torn down, rotted away in wasting heaps of fallen leaves. Here and there, some of the thick rind of the house had peeled off in strips, and fluttered heavily down, littering the street; but, still, below these rents and gashes, layers of decomposing posters showed themselves, as if they were interminable. I thought the building could never even be pulled down, but in one adhesive heap of rottenness and poster.
This landscape is the result of the intense competitiveness of bill sticking, which was not effectively regulated. The outcome was a universal blighting of the environment, as vividly portrayed by E.S. Turner in The Shocking History Of Advertising:
In the [eighteen] thirties and forties London was papered over, nightly and often twice-nightly, until it vanished from view. […] Every vacant wall was pasted and overpasted with shrieking announcements barbarously printed. […] The real problem for the ‘external paper hanger’ was to operate early enough in the day to escape interference, yet late enough in the day to be able to cover a rival’s posters. Contractors banded together some of these guerillas, with the result that pitched battles occurred from time to time between rival gangs. Since the fly poster’s stock-in-trade consisted not only of paste but of blacking to deface the enemy’s handiwork, these encounters were as spirited as anything in a Keystone comedy.
In this milieu, speech and counterspeech physically accrete on any available surface: layers upon layers of rival sloganry encrust the townscape and beyond. This is literally a war of words, a battle for square footage, with every available surface bearing the visible evidence of skirmish.
Turner’s book is a gem, covering the little-considered backstory of how advertising speech evolved in different media — along with the mechanisms for its control (in the UK). The ever-increasing nuisance of advertising as techniques evolved led to the imposition of restricting laws, amid lobbying from groups like SCAPA (The Society for the Checking of Abuses in Public Advertising).
Public opinion against advertisers was driven by their excesses. Grandiose, ever wilder schemes to interpolate advertising materials into every conceivable plane of visual space flourished in Victorian times. Sites of beauty were despoiled, leading to hoardings on the Great Pyramid at Giza, billboards on the white cliffs of Dover, and a host of other aesthetic insults:
A Gothenburg margarine factory had painted its slogans on the walls of the fjords, right up into the Arctic Circle. The rock faces of the Thousand Isles of the St Lawrence were hideously daubed with slogans for stove polish and tooth powder. Out of the Mississippi rose letters twenty feet high testifying to the virtues of a chewing tobacco. […] Every other rock on the canyons of Nevada was disfigured. In one of the loveliest bends of the River Reuss in Switzerland a giant boulder had been painted brown, with a chocolate firm’s name on it. High on the Rigi were enormous gilt letters advertising a hotel. Even the remote hills of the Sudan carried slogans for soap.
Technological innovations continued to create novel kinds of encroachment: at one point London was plagued with ubiquitous sky signs, towering letterforms erected on the roofs of buildings; the unrestrained use of flashing electrical signs transformed the nightscape; and there were various different trials of the projection of advertising onto the landscape and clouds.
It’s easy to see that the correct response to companies projecting their brand slogans onto the sky is not to bleach the night with counterprojections. The correct response is to stop the defacement of the sky — to curtail the visual pollution. Unshackled, advertising rapidly expands to fill all space available: legislation (“enforced silence”) is ultimately the only effective limiter.
In the case of advertising, the best counterspeech is not a proposition, but an imperative: “POST NO BILLS”.
Bankruptcy In The Marketplace Of Ideas
It’s ironic that the bromide of “the marketplace of ideas” is used to cheerlead for the counterspeech doctrine, while advertising — the speech of the marketplace itself — requires close regulation.
In fact, the despoiling effects of market operation require not just advertising regulation but food purity law, environmental law, competition law, and so on. In reality, “the marketplace” is a terrible metaphor for the idea that speech competing with counterspeech would lead naturally to the optimal outcome of a discovery of truth.
Online discourse is particularly illustrative of bankruptcies in the marketplace of ideas. While history is strewn with examples of irrational movements and happenings (Extraordinary Popular Delusions And The Madness of Crowds gives a sampler), the speed, frequency and intensity with which such ideas have taken hold in the 21st century are startling: the internet appears to be an accelerant for the wildfire spread of quackery.
Information may want to be free, but there is no corresponding compulsion for it to be accurate.
We are bombarded online with the deluded ruminations of bigots, of anti-maskers and anti-vaxxers, and of manifold other conspiracy theorists with their baroque confabulations around topics like 9/11, mass shootings, 5G phone signals, and election rigging. Descenders of online rabbitholes may even outright repudiate basic scientific knowledge, as seen in revivals of germ theory denialism and flat earth beliefs. And multiple strands of such discourse are melded in grand convergences of unreasonableness, phantasmal cathedrals of contrarian confabulation such as QAnon. Traction for discourse about “lizard people” suggests no notion can be too outre or too shoddily scaffolded by evidence to gain adherents.
The widespread embracing of these ludicrous notions seems to be simply due to the ease with which online platforms allow for raw assertion. As with advertising, the ubiquitous redistribution of assertions is sufficient to perform the function of creating “brand awareness”, which supercedes any need for evidence; such assertions may even thrive in the face of contra-evidence. Their incessant reiteration can be used to define the landscape of discourse, provide a drumbeat for marching to, and can simply overwhelm counterspeech by dint of sheer repetition: “no smoke without fire” is too easy to assume.
The ease of reiterating assertions in online platforms through speech acts such as reshare or like means this technique of propaganda is baked into their content creation and distribution mechanisms: approbation equals redistribution, and redistribution is the basis of monetization. This works to drive the creation of content which works void of context, as a redistributable atom of assertion, a mosaic tile.
The gaming of platforms makes matters worse. The terminology of online speech alone reveals a rich and complex ecosystem of novel speech acts — dogpiling, sealioning, sock puppeting, gish galloping —that are used to suppress, overwhelm or submerge counterspeech. (These are “mob speech” techniques that would not be possible in a moderated face-to-face discussion, a toxic combination of mendacious/loquacious behaviours).
Online, we are amidst virtual bill sticking wars, viewing a chaotic wasteland of meaning like that of Victorian advertising; a jumble of argumentation.
Diegetic Criticism
Some limited attempts have been made at introducing counter-assertion mechanisms, like fact-checks, to online platforms — but simply adding an anodyne warning label to a piece of content that can be endlessly reasserted in different variants seems to be hopelessly feeble; who reads past the headlines anyway?
Perhaps the existence of ambivalent caveats is an improvement on zero pushback, but this doesn’t effectively tackle the environmental degradation that endless reassertion creates. Without more robust pushback, media platforms will come to resemble Dickens’ description of a bill-stricken building:
an old warehouse which rotting paste and rotting paper had brought down to the condition of an old cheese
A simple pointer towards methods of “uncheesing” the informational landscape is Shinigami Eyes, a browser plugin that works as a crowdsourced reputational indicator. When installed, content that originates from transphobic people or sites is shown in red, and content from trans-positive sources in green.
This immediately orients us to the attitudes and histories of participants within a discussion. It allows us to understand who has a history of bad faith, prejudiced or conspiracy theoretic speech (red), who is generally trustworthy (green), and who (uncoloured) might be an innocent bystander trying to make sense of the hostile cross-fire they are caught in. At a glance we can see what is likely to be worth paying attention to, and to where counterspeech might be effectively directed.
Shinigami Eyes enhances our understanding of online conversation not through counterspeech but through counterintelligence. It layers tonal information onto content, just as pitch or rhythm in vocal speech can be used to signal emotion/attitude. This is an example of diegetic criticism: by which I mean context/exposition on a narrative or discourse being interwoven into the content through the use of additional modalities or layered information, similar to the use of prosody in spoken language, or narration or extradiegetic music in a movie.
Diegetic criticism differs radically from standard platform mechanisms for signalling approbation, such as “likes”, which have no impact beyond the fragment of content they adhere to. It is, instead, tied to a source of content: it maps from the overall behaviour of content creators to an impact on their overall content, on an ongoing and dynamic basis. All of the content ever created by that source exposes the reputation of that source, by displaying a trait immediately apparent and understandable.
(Problematical content is sometimes flagged by platforms: for example, automated detection of “abusive language” might push content into a click-to-reveal container. But this is tied to textual analysis of specific pieces of content, not on crowdsourced insight into traits of the content author.)
We can imagine a variety of other interface tweaks to use for the amplification or diminuition of content — changes in font size, color saturation, content reordering or segregation, and so on; some tweaks might perhaps be gradated in intensity depending on the constancy of awfulness or brilliance that a content source exhibits.
This kind of modification of the virtual environment allows for perceptions of discussion to be rejigged. Ideally, we would be able to use this approach within any platforms to highlight or amplify reputable sources, and to treat sources of bigotry, abuse or misinformation with appropriate suspicion. But unfortunately tools like Shinigami Eyes are limited to what can be achieved outwith apps, as a browser plugin.
If platforms allowed explicitly for the embedding of diegetic criticism, it would allow us to expose “falsehood and fallacies” through community refereeing, to provide a means of visible shaming (insofar as it is possible to shame the shameless) powered by “many eyeballs” scrutiny. Unlike the despoilation of the physical environment by advertising, the environmental degradation of virtual space which endless reassertion creates could be tackled by patching the environment itself, without the need to “silence” awful ideas through regulation.
The ultimate question for all reputational mechanisms is whether they can be built in a way that is resistant to gaming: whether their mechanisms (and therefore their own reputation) are resistant to attack.
At a small scale, a tool like Shinigami Eyes can be monitored and maintained by a single motivated curator. At a larger scale — a platform like Stack Overflow, for example — reputation can itself be tied to moderation privileges: a valued contributor to the system is allocated juridical powers, allowing attempts to derail the system to be counteracted. Shared responsibility and diverse perspectives provide a means of hedging against partisan refereeing.
Coming Unstuck
Assuming that the autocorrection of a “marketplace of ideas” applies within today’s media landscapes is foolish; we can see many patently bad ideas taking firm hold of crowds in near real-time. A Panglossian assumption that the best ideas will inevitably triumph does not serve us well: there is little chance that errors of fact or reasoning can be corrected simply through argumentation in environments that are subject to partisan refereeing.
Online platforms, while superficially appearing to enable social deliberation, are actually a source of novel speech acts that allow for the incessant reiteration of assertions. The self-refereeing of those platforms is frequently ineffectual or erroneous; they often fail to correctly enforce their own stated policies against extreme and harmful content, for example.
Some of the characteristics of online platforms that militate against balanced refereeing are:
- Platforms have limited motivation to act upon anything but engagement (which is the basis of monetization)
- Conventional kinds of refereeing mechanisms are simply unable to handle the firehose speed of content generation and the co-ordinated deployment of mob speech techniques
- The refereeing that does exist does not catch dog whistling: toxic content which is carefully tailored to allow radicalizing messages to be disavowed, by disingeniously pointing to their surface level innocuousness
So, sometimes the best response to speech that is possible may be simply to provide enhanced scrutiny of who is engaging in conversation. If we can’t|shouldn’t stop a flow of foolishness, we can at least turn to signal jamming.
We need tools that assist the cultivation of an appropriate disdain for untethered assertions, the stuck bills of our modern day purveyors of bullshit.
Allowing the agenda and alliances of a speaker to be uncovered through diegetic criticism gives us a means to weigh trustworthiness, to judge the value of words. The judgment of topic experts can be shared with the less intensely engaged, allowing for clarity about the motivations that are at play within a dialogue — or a monologue masquerading as one.
Bonus track:
