The French theorist who predicted ChatGPT in 1967

Here’s how a human author writes, according to French literary theorist Roland Barthes in 1967:

“The text is a multi-dimensional space in which a variety of writings, none of them original, blend and clash. The text is a tissue of quotations drawn from the innumerable centers of culture… The writer can only imitate a gesture…his only power is to mix writings, to counter the ones with the others.”

If you think ChatGPT isn’t creative, then Barthes would respond, neither is the human author. According to Barthes, the author never writes anything original. The “author” as a solitary genius is a myth. The mythical “Author” has been replaced by a scriptor:

“The Author is thought to nourish the book, which is to say that he exists before it, thinks, suffers, lives for it…in complete contrast, the modern scriptor is born simultaneously with the text, is in no way equipped with a being preceding or exceeding the writing.”

Roland Barthes, 1967, “The death of the author,” Aspen Magazine, 5-6, section 3.

With ChatGPT, you can say “write an essay about Marx’s theory of value” and ChatGPT will do an amazingly good job. It does it by absorbing every single text, every sentence, that is available on the internet. It puts those gazillion sentences into a big blender, and it spits out a new text assembled from those bits and pieces of literary culture. It’s both original and not original. If you wrote your dissertation on Marx’s theory of value in 1978 or 2011, then ChatGPT probably stole a tiny little percentage of it. Intellectual property scholars are having a field day exploring the implications. Can you sue ChatGPT for plagiarizing your dissertation? How about a class action lawsuit by all of those uncited and forgotten Marxist dissertation writers over the past 50 years? Maybe you’ll recognize one of your own sentences, from page 136 of your thesis.

Barthes’s essay was widely cited and became gospel among post-structuralist, post-Marxist literary theorists through the 1970s, 1980s, and 1990s (see my book Explaining Creativity, 2012, pp. 333-334). Marxist theorists argue that the self is socially constructed and there is no true individuality. The author is nothing more than a mediator, a curator, through which the massive texts of culture flow. If these theorists were right, then ChatGPT creates in exactly the same way as a human author–sorry, I mean “scriptor.”

Here’s what ChatGPT itself says. I typed in the prompt: “How is Barthes theory of the author as scriptor similar to generative AI?”

“Barthes theory of the author as scriptor views authorship as a process of assembly, whereby authors bring together existing elements to create something new. This idea is similar to generative AI, which uses existing data sets to create something new or novel. In both cases, existing elements are used to create something that does not exist in its entirety before the process begins.”

4 thoughts on “The French theorist who predicted ChatGPT in 1967

    1. Thanks Johnnie! That group of 1970s French scholars became incredibly trendy in intellectual circles of literary theories and writers. That’s not my own scholarly home, but it has obvious implications for creativity theory. If the author is essentially a hub in a cultural network, a hub that connects many existing ideas that are part of the culture, then it changes how we think about creativity. Creativity becomes a culturally-embedded act of linking shared distributed conceptual networks. Creativity is no longer an inner mental state that can be fully understood by psychological experiments. I call this the “sociocultural” view of creativity in my 2012 book and it’s the approach I take in my own research.

  1. Hello Keith – Of all the so-called technical “improvements” to our lives that are anything but – don’t start me on social media! – this is by far the scariest. Once it becomes widely accessible, its implications for education are staggering. It also further devalues the individual voice, which means dismissing the individual thought and diminishes the singularity of creativity. Orwell would have a field day with this stuff!

    1. The issues aren’t really that new, either. In the 2006 first edition of my book EXPLAINING CREATIVITY I had a chapter on computational creativity, with reviews of early programs like AARON (Harold Cohen’s AI painting program), story-writing programs including TALE-SPIN and MINSTREL, and EMI (David Cope’s “experiments in musical intelligence”).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s