5: Circulation

History of Circulation 

Aristotle suggested that to be an effective writer or speaker, a student must become skilled in five different aspects of communication: invention (generating ideas), arrangement (organization and order), style (words and grammar), memory (memorization), and delivery (speaking or writing). Aristotle argued that effective delivery was just as important to communication as the ideas themselves or the language used to convey them.  

In ancient Greece, most ideas were communicated orally, through public speeches, debates, or conversations. For Aristotle, the concept of delivery meant giving speeches, participating in debates, or having conversations. Most simply, the concept of delivery has always focused on how an idea is shared from one person to another, between a speaker and audience, or an author and readers. Speech and performance were the dominant modes of communication in the ancient world, and so it followed that discussions of delivery focused on the spoken voice and bodily movement. Only with the development of bound and printed texts like books, newspapers, and posters did the ancient concept of delivery evolve to focus more on the written word. 

Recording ideas in handwritten texts was one of the first steps in allowing ideas to circulate in new ways. Making copies of texts or manuscripts required scribes to copy by hand one word at a time onto blank pages to create a new manuscript. Because these texts or manuscripts were rare, most manuscripts produced by scribes ended up in the libraries of monasteries or wealthy individuals.  

Therefore, one might understand why Gutenberg’s invention of the printing press in 1455 was one of the most important moments in the history of the circulation of the written word. His invention meant that more books could be produced. As a result, books became more plentiful and inexpensive. More people than ever before had access to the world of ideas through the written word.  

As the transition from quill to printing press illustrates, available technologies can impact circulation of a text: from how it’s produced, to how it’s consumed, to what its material properties are, and who has access to it. 

Circulation Today 

This sort of transition, from one communication medium to another, has become increasingly common as new technologies emerge. The invention of recorded sound and radio radically shifted the way we produced and consumed texts. Certain types of texts were no longer limited by space and time. A symphony performed by an orchestra in London could be experienced by a banker in Cleveland months or years after the actual performance. The same can be said of motion picture cinema and broadcast television. Moreover, the development of radio stations, movie studios, and television networks has always been followed by amateur producers who take advantage of consumer technologies like ham radio, photocopiers, video cameras, or cassette tapes. All of these technological developments have produced new types of texts and new media.  

Further, as our writing and media technologies have become more powerful, our texts increasingly seem to take on a life of their own. This idea has never been more current than in our contemporary digital culture of websites, social networks, viral videos, and internet memes. There used to be a clear distinction between producing a text and distributing it. Writing a text and publishing a text were two different activities entirely. The same could be said for recording a song and playing it on the radio or making a movie and showing it in a theater. Writing and sending a letter was a process measured in days, not seconds. Making a movie required a production studio, rather than a smartphone. New technologies continue to emerge at an increasingly faster rate, and the distinction between producing a text and circulating it is slowly disappearing. As a result, writers who have the best understanding of how texts are shared between people, how they circulate within and across social networks, will have the most impact and the biggest presence in the networks most important to them. 

Wikipedia is an excellent example of the way that networked and digital technologies have changed the very nature of how certain texts are produced, circulated, and evolved in contemporary culture. For centuries before the internet enabled and democratized the digital publication and distribution of texts, encyclopedias were one of the most important and respected repositories of knowledge and reference information in almost any home or library. An encyclopedia was usually made up of dozens of expensive bound print volumes/books. Usually, each volume was dedicated to a letter of the alphabet. Unless a researcher had access to the encyclopedia’s most recent edition, the information contained within its covers was likely not current. Even the most recent edition was likely written months or even years earlier due to the lengthy process of editing and publishing bound books at that time. 

These limitations were taken for granted in an age where print-publication dominated the circulation of texts. But there was no alternative. Acquiring a new, yearly volume that was produced every year to update a set of encyclopedias was often a financial impossibility for most families and libraries. 

As new technologies became available, Wikipedia was designed specifically to respond to these exact limitations, including restraints of accessibility, up to date information, and cost to name a few. Although not a perfect solution to the challenges noted above—there are some who question the veracity of some entries within it—Wikipedia is accessible to any individual with online access, whether that be at home, at school, or at a local library. But accessibility is only one of the elements of circulation that Wikipedia addresses. Each of the entries within Wikipedia is constantly evolving. As new knowledge becomes public, as science charges forward, and political events unfold across the globe, Wikipedia is able to accommodate and incorporate those bits of knowledge far more quickly and (we hope) democratically and (we hope) accurately than ever before. Since the advent of Web 2.0 technologies, readers of Wikipedia have been able to participate in its existence as writers and collaborators. They can suggest corrections to inaccuracies or add emerging knowledge to the existing entries, and their suggestions become available almost instantly. Teams of volunteer editors monitor pages in their subject areas, enforcing style and attribution guidelines, sometimes controversially. Nonetheless, occasionally errors find their way onto pages for hours, weeks, days, or even longer. This interaction, where the boundaries between reader and writer, producer and consumer begin to blur, has resulted in one of the most comprehensive and highly accurate sources of knowledge ever produced.  

A fascinating example of just how powerful the idea of circulation can be is the internet phenomenon of fake news, which—though it existed well before then—came to light during the 2016 presidential election. Keep in mind that authors can make money based on the number of clicks received for an article they have written, so a story that has a high circulation and gets attention can mean more money for an author even if it is fake. However, some fake news stories are intended to sway public opinion about particular issues or people. In both cases, those that are written just to circulate and those written to influence people’s beliefs are deliberately written to contain misinformation. These are hoaxes, not satire: Satire calls attention to itself as untrue, whereas fake news tries to convince readers that it is true. 

“Successful” fake news authors have mastered the art of circulation: they think about who their readers are and produce easily shareable content, anticipating that their readers will spread the news via social media. One fake news story in late 2016 spread the lie that a pizza parlor in Washington D.C. was the site of a child sex slave ring led by Hillary Clinton. This story motivated a man to show up at the pizza parlor with guns, allegedly to free the children (Kang and Goldman). The power of the written word in the age of the internet is profound due to the increased means of circulating what we write, but the story of the pizza parlor is a reminder that circulation without considerations of ethos and ethics can be truly dangerous, too. 

Alphabet (parent company of Google) and Meta (parent company of Facebook) have taken some steps to change their algorithms to reduce the visibility of fake news stories, but the danger of being fooled is still present for anyone who spends time on the internet.  

Circulation and Other Concepts 

It’s one thing to think about circulation in terms of viral videos, internet memes, or popular movies and television, but it’s just as important (and more practical) to think about circulation in terms of texts being exchanged among a smaller group of people who share a common field. For instance, a scholar might write an article for a journal, and that person could reasonably assume that it would only circulate among other scholars interested in that topic. Scholars have a sense of the people who might encounter their texts, so they can make intelligent guesses about their audience’s expectations regarding the features of the genres they produce: citation styles, referencing the work of previous scholars, or using certain words that might only be familiar to the people among whom the texts circulate. 

Certain types of text afford different types of circulation. It’s not likely that you’ll see a brilliant meeting memo being shared among friends, and you likely know better than to include a link to your favorite cat videos as part of your job application materials. If you want something to be shared among friends, you’re more likely to produce an image with a caption or a video, maybe even an Instagram post. If you’re responding to a request for job application materials, you’ll write something meant for one or two specific people, and you won’t have to worry about producing a text that could be easily shared outside the company to whom you are applying.  

When looking at circulation and affordances, it’s best to think about these as being dependent on each other. If you want something to circulate to as wide an audience as possible, there are certain choices you can make about the modes and genres you take advantage of to produce your text. If you want to be well-known or an influencer and respected for being able to do tricks on your bike, you’ll likely put together a demo or highlight video and post it to YouTube and other sources so it can be shared among as many people as possible. On the other hand, if you know that you’re only going to have access to email or a text message to communicate your idea, you can guess (or hope) that it won’t circulate much beyond the people to whom you send it directly.  

Sometimes circulation will be limited or determined by the affordances of the materials to which you have access. But at other times, you might have a very specific sense of the audience among which you’d like your text to circulate, even if it’s just one person, and you’ll select your materials accordingly. 

Looking Ahead 

One of the aspects of studying any conversation is understanding how it circulates. We know that circulation refers to how texts move from authors to readers and how texts are shared within and across different communities. When we immerse ourselves in an intellectual conversation, our effort to listen carefully reveals the important voices and the ways ideas circulate among participants in the conversation. We find ourselves focusing on the contributors who demand our attention through the insight and clarity of their ideas as well as their ability to access and dominate the networks, platforms, and genres where their work circulates. Think for a moment about the students of Stoneman Douglas High School and their success at puncturing through the rhetoric of the gun lobby. First, those students had to survey the many perspectives that make up the conversation on gun control and then decide how they could create a message that would move the conversation forward. Their mastery of social media platforms enabled them to create effective messages that spread rapidly and built momentum. As their messages circulated, other young people were able to navigate and invent their own messages, kairotically surfing and selecting from all the available media. We write to circulate, but we also write with, from, and because of circulation. 

As you enter conversations and contemplate what your contribution will be, consider how the concept of circulation can help you at every stage of your work: when you are gathering and reading sources, when you are planning what your contribution will be, and when you are imagining how your message will spread. 

Circulation in AI-Assisted Writing 

AI technologies based on large language models are already shifting the conversation around circulation in at least two broad areas. First and perhaps most obviously, AI shapes content being circulated through existing systems for sharing information, such as popular social media platforms. Various AI language models have been trained to fabricate a wide range of content: written text, certainly, but also software programs, images, spoken audio, music, and video. A number of these technologies are designed to assist people in the work of content creation. For example, in September, 2023, Google announced a suite of AI-powered tools to translate spoken audio into other languages, generate outlines for videos that YouTube creators can make themselves, and produce background videos and images that can be incorporated into YouTube Shorts (Sato). These tools offer clear benefits to users, helping them develop richer content that can reach a wider audience in less time. For those who earn an income from YouTube, the ability to generate more robust videos more quickly can translate directly to economic gain. For more casual or even first-time posters to YouTube, these tools can have a democratizing influence, making the idea of becoming a content creator on the platform seem within reach and encouraging new voices to circulate on the platform.  

These benefits do not come without additional challenges. In the same month that Google revealed its AI tools to support content creation, Amazon announced two changes to its policy governing the self-publication of Kindle e-books on its site. First, Amazon added a new requirement that authors notify the company if their e-books contain AI-generated material. The policy focuses on AI-generated content specifically. Authors who rely on AI tools to assist in their own development of content do not have to identify their texts as products of AI. Interestingly, Amazon did not commit to passing this information on to consumers: the company would know which products available in its store contained content created by AI, but purchasers would not (Amazon indicated it might alter its disclosure policy in the future to make this information public) (Italie). A matter of days later, Amazon followed up with a second update to the policy, restricting authors to publishing no more than three new e-book titles per day on the platform (Creamer). As the company explained on its Kindle Direct Publishing forum, Amazon “is actively monitoring the rapid evolution of generative AI and the impact it is having on reading, writing, and publishing” and is “committed to providing the best possible shopping, reading, and publishing experience for our authors and customers” (“Update”). Taken together, these policy changes suggest a number of ethical, economic, and legal questions. Why is it important for Amazon to know which titles are written by AI models but not for consumers to have the same information? What assumptions about human-generated and AI-generated content frame the claim that knowing the authorship of Kindle books will enable Amazon to provide the best experience for writers and consumers? How do authors, readers, and Amazon factor AI into calculations of value, understood as both “quality” (however one cares to define that concept) and price (a fair and reasonable cost that compensates authors for their labor and Amazon for their services while providing consumers with the content they want)?  

Consider some of your own perspectives on these questions. If you are a content creator on the Kindle platform or a similar service, or if you envision yourself possibly pursuing these opportunities in the future, how might you feel about earning the same compensation for your labor as an AI that generated competing texts in a matter of seconds? As a consumer of content, would you want to know whether the content you purchased was written by people or by an AI? Why or why not? Would you believe it appropriate for you to pay the same price for similar products regardless of how they were developed? How much of your sense of the fair value of a product comes from your perception of how useful the product is to you, and how much might fair value reflect your understanding of the amount of labor that went into the generation of the product? All of these questions stem from the pressures that AI sparks in our current systems of content circulation and in our attitudes toward and perceptions of how content should circulate based on shifting historical and technological conditions.  

If content is the first dimension of circulation affected by large language models, a second area might be the technologies of circulation themselves. AI models have implications for how content—regardless of how it was developed—circulates. A September, 2023 article (the prominence of this particular month in this discussion is one indication of how rapidly new changes and questions resulting from developments in AI models are emerging) on the science and technology news site Ars Technica highlights ways that AI might improve data compression. While fascinating, the technical details are less relevant here than the ramifications for data transfer. We rely upon data compression all the time: phones, streaming services, still image and video cameras, the Internet, satellite communications, and nearly all of the other information infrastructure we routinely interact with make use of data compression algorithms to squeeze as much full quality or “good enough” information into as little bandwidth as possible. The more efficient the compression, the more data can be circulated at less cost. The Ars Technica article explores a recent study–yes, from September 2023–that demonstrates the ability of large language models to compress some data types more efficiently (sometimes significantly more so) than commonly-used compression algorithms such as PNG for images and FLAC for audio without any loss in quality (Edwards). AI thus has the potential to streamline the transfer of extremely large data sets that are costly and lengthy to circulate using current algorithms. It might lead to reduced costs for data transfer and possibly to reduced energy usage, resulting in a smaller environmental footprint for digital data circulation. In time, these changes could alter authors’ sense of the resources available to them and of the size, scope, and format of the content they circulate. 

Writing with Circulation 

With so many texts competing for our attention and our memory, it’s important to understand how texts operate in people’s everyday lives. Knowing how receivers might encounter a text, or when it might be most appropriate for them to receive it, will often make the difference between being noticed and being passed over. Some things to consider are: 

• What sorts of things make a text memorable? Why do I notice or remember certain texts and not others? 

• How are texts circulated in my particular field? What genres circulate most? 

• In what sorts of social networks do the people I’m trying to reach participate most? Do people participate more as a consumer, producer, commenter, or sharer in these networks? Do they share things from one network or platform to another? Why? 

• What sorts of experiences, good or bad, have I had with the means of publication on the web that focus primarily on alphabetic text (Google Docs, blogs, Tumblr, wikis, forums, etc.)? Which ones do I prefer, and which do I think my audience will prefer, and why? What is it about them that attracts me to them? 

• What are the affordances or constraints of the various methods available to me for circulating my text? 

• What image have I created for myself? Am I accurate, reliable? Do I understand my intended audience? What have I done to indicate these? 

Works Cited 

Aristotle, and George Alexander Kennedy. On Rhetoric : A Theory of Civic Discourse. 2nd ed. Oxford University Press, 2007. 

Creamer, Ella. “Amazon Restricts Authors from Self-Publishing More than Three Books a Day after AI Concerns.” The Guardian, 20 September 2023, Amazon Restricts Aughots [theguardian.com]. 

Edwards, Benj. “AI Language Models Can Exceed PNG and FLAC in Lossless Compression, Says Study.” Ars Technica, 28 September 2023, AI Language Models [ arstechnica.com]. 

International Federation of Library Associations and Institutions (IFLA). How To Spot Fake News. 2017. How to Sport Fake News [ repository.ifla.org]. Licensed under CC BY 4.0.  

Italie, Hillel. “Amazon to Require Some Authors to Disclose the Use of AI Material.” Associated Press, 8 September 2023, Amazon to Disclose the Use of AI [ apnews.com]. 

Kang, Cecilia, and Adam Goldman. “In Washington Pizzeria Attack, Fake News Brought Real Guns.” New York Times, 5 December, 2016, Fake News Brought Real Guns [nytimes.com]. 

Sato, Mia. “YouTube is Going All in on AI.” The Verge, 21 September 2023, YouTube is Going All in on AI [theverge.com]. 

“Update on KDP Title Creation Limits.” Kindle Direct Publishing Forum, 18 September 2023, Update on KDP Title Creation Limits [kdpcommunity.com]. 

Wikipedia. Wikipedia Logo. 2011. Widipedia Logo [ en.wikipedia.org]. Licensed under CC BY-SA 3.0.  

Images

Circulation Image - Wikipedia .docx