“NRA Tally (@NRA_Tally)” by Mark Sample

Tweets Following Followers 159 0 24 NRA Tally @NRA_Tally Keeping score of the NRA's greatest hits. Fairfax, Virginia everyadage Kathi Inman Berens Brett O'Connor Alex Gil Followed by everyadage, Kathi Inman Berens, Brett O'Connor and 2 others. NRA Tally ‏@NRA_Tally 38m 30 postal workers killed in San Francisco with a AR-15 assault rifle. The NRA steps up lobbying efforts. Details NRA Tally ‏@NRA_Tally 4h 22 restaurant diners murdered in Jacksonville with a 10mm Glock. The NRA reports a fivefold increase in membership.
Open “NRA Tally (@NRA_Tally)” by Mark Sample

Created in the wake of a mass shooting event in Isla Vista, California, this bot takes aim at the National Rifle Association and the rhetorical strategies it uses to protect the industry and gun culture it lobbies for. He accompanied it with a manifesto titled “A protest bot is a bot so specific you can’t mistake it for bullshit: A call for bots of conviction” in which he invites the creation of bots which are “topical, data-based, cumulative, and oppositional” (here’s an updated version). He also explains how his bot @NRA_Tally meets these characteristics and goes into great detail on the data sources that inform the bot’s generation of murderous hypothetical scenarios, such as this one:

It is not difficult to intuit the variables from reading multiple tweets in the bot’s Twitter stream– a number, victims, location, firearm, and response– but I encourage you to read Sample’s excellent essay which explains in detail the real and abstracted data sources that inform this bot. His selection of data sources and rationale for each range of possible values enrich our understanding of this bot beyond what can be determined solely by reading its output, to the extent that the essay could be considered a part of the work itself.

So what is the effect of following this bot on Twitter? Why would someone want to sprinkle horrifying fake shooting spree scenarios and cynical NRA responses into their Twitter stream? In a word: inoculation. When you are exposed to regular small doses of this toxic juxtaposition, your mind begins develop resistance to the NRA’s rhetorical strategies. A potential side effect of this treatment is desensitization to mass shooting events, but let’s face it, the frequency of real mass shootings in the United States is enough to desensitize a whole country to the point of inaction, a point made eloquently by The Onion with its recent satirical work “No Way to Prevent This.”

The poetic interest of this protest bot and the computational Onion article is in how they focus what I call the e-poetic function of language. This is an extension of Roman Jakobson’s notion of the poetic function of language, one of six functions of language he described in his seminal “Closing Statement: Linguistics and poetics” in 1960. The poetic function of language is when language draws attention to the message for its own sake, and while it is the defining function of poetry, it is used in other fields, such as advertising and slogans. In practice, Jacobson effectively used linguistics to analyze the phonetic, syntactic, and semantic qualities of language, with a focus on the two primary poetic media for the time: writing (in print), and orality (performative and recorded).

The e-poetic function of language extends Jacobson’s poetic function by drawing attention to the intersection between language and electronic/digital media, which adds programmed behaviors to the language (see the I ♥ E-Poetry Metadata page for a detailed explanation)– that is, what the words do and under what conditions. For example, @NRA_Tally is a bot that generates static content every 6 hours, and these are constituitive components of its poetic function. The work is not contained in a single tweet, or even in all its current tweets: it is in the generative algorithms expressed by the programming code, in the very rhetoric of its code and variables, in its data sources and data sets, in the platform (Twitter) through which it is deployed. This bot operates with strong referential and emotive functions of language, but its dominant function is poetic (see Waugh)– or e-poetic, in my formulation.

And just as Jacobson examines the poetic function in fields other than poetry, I seek to explore the e-poetic function in other uses of language in electronic & digital media, particularly when used for persuasive reasons, identifying these cases with the e-rhetoric category.

So visit or follow @NRA_Tally, thinking of how its generative scheduled tweets, establish a rhythmic and persuasive poetic experience through repetition, insistence, and emphasis. And don’t overlook how it uses location in its Twitter profile as a targeting device for its readers’ outrage.

Works Cited

Flores, Leonardo. “Digital Textuality and its Behaviors.” On Intermedial Aesthetics and World Literatures (2013): 123-139.
Jakobson, Roman. “Closing statement: Linguistics and poetics.” Style in language 350 (1960): 377.
Waugh, Linda R. “The poetic function in the theory of Roman Jakobson.” Poetics Today (1980): 57-82.