Silly Elephants Dancing Without Meaning? On "Paratexts Without Texts"

Curator's Note

I will never see this film. It should therefore seemingly only ever be a “paratext without a text” for me. And yet it is a text. This trailer brings me pleasure in its sheer awfulness, and creates a text that has meaning for me. Thus, as important as I think it is to study paratexts when what Barthes would call their accompanying “work” is absent, I offer this as an example of how paratextuality is inseparable from the text. If the paratext means something – anything – it is part of the text. And we can’t truly appreciate that meaning, nor paratexts’ role in the construction of meaning in general, if we see the paratext as removed from the text.

Here, I diverge from Gerard Genette, who offered us the term “paratext,” but forestalls its utility when he insists of “paratexts without texts” that, “The paratext is only an assistant, only an accessory of the text. And if the text without its paratext is sometimes like an elephant without a mahout, a power disabled, the paratext without its text is a mahout without an elephant, a silly show” (410).

This is a very bad metaphor (“C Me Metaphor”?). Most obviously, its desire for Orientalist spectacle and its devaluing of an Indian man without his elephant is cringeworthy. But it also doesn’t allow for a paratext to create the text, to be the text. C Me Dance’s trailer is definitely a silly show, but not because it needs an equally silly elephant. The trailer effectively conjures up the text, or as much of the text as I will ever need, want, or consume. It needs no accompaniment, and is far more than an “accessory” for me as viewer. In Barthes’ terms, I have no familiarity whatsoever with the work, but there most definitely is a text, constructed in total for me by this trailer. Paratexts are always already part of the text, so one simply cannot have a paratext without a text, no matter how much this ticks off the Devil.


Johannes Mahlknecht's picture


Hello Jonathan, and thanks for your post. It really reminds me of what Vincenz Hediger once said at a conference in Innsbruck about trailers in general; that they’re so great because they “spare you the trouble of watching the whole movie” (I’m sure I’m misquoting here, but that was the gist of it). His case in point, I think, was “Rambo: First Blood Part II” (why not just “Rambo: Second Blood”?), which by the way I still think isn’t quite as terrible as everybody says it is. I any case I believe that especially within the realm of Hollywood entertainment, where generic streamlining of products is still very much the done thing, trailers are wonderful tools to separate the wheat from the chaff, so to speak. Even if I want to think that I’m not one of those superficial persons who judge a book by its cover (or film by its trailer) - how often do I find myself thinking after watching only half a trailer: “Okay, seen that movie, next please!”? So what I’m particularly interested in in connection to paratexts in film is to observe how much text paratexts can actually contain, and how much information they can transport in combination (say, poster taglines + poster image + 3 or more trailers + promo clip = the movie?). I wouldn’t say that the paratext IS the text (although in spoof trailers such as my “Minesweeper” example that idea is kind of the point, and I agree that in your example it definitely SHOULD be), but with ever more and ever longer trailers for individual films being released, we may well be getting there…

Jonathan Gray's picture


Thanks Johannes. I love the idea of seeing how much a trailer can carry (more than an elephant, dare I ask?). And what they can’t carry. This all makes me think too of the wonderful Onion News Network piece about the planned adaptation of the Iron Man trailer into a full, feature-length film

Ellen McCracken's picture

Are we using "text" too broadly?

Love your work, Jonathan, but like Johannes, I’d urge a bit more refining of categories here. It’s great that you’re pushing us to reconfigure what we mean by “text,” and provocatively suggesting that text and paratext are one. But can we argue, for example, that students who read only the reviews know the literary text? Yes, those short takes constitute THEIR textual knowledge, but don’t we have to distinguish varieties of our overarching term “text” in these cases? (I like that you at least call it ‘a’ text at the end.) But after all, your own careful reading of Genette revealed the “Orientalist spectacle” that you might not know about through internet summaries or reviews—that primary text is necessary. Even if we hate paratexts like the C Me Dance trailer, don’t we need someone to do the work of analyzing the whole messy text to be accurate about what’s going on there? Yes, the trailer “creates a text that has meaning for [you],” but can this be conflated with the movie text? You, as a poaching reader, say “basta” after viewing the trailer; great example of how readers in fact create texts. But we still need a term for the species of text that is the movie—that for you is a text not actualized, a potentiality.

I agree that some paratexts are not really paratexts, so linked are they to the primary text. I’ve argued this about Junot Diaz’s footnotes in The Brief Wondrous Life of Oscar Wao, despite their position at the bottom of the page. Here I would agree, as you say, the paratext has lost its function as a paratext but now this has occurred because of substitution—it was enough for you. But trailers, although strongly linked rhetorically and on multiple levels to films, are still separate, albeit interlinked entities.

So let’s work on some subcategories of the term “text”to analyze our way around this minefield, while still keeping your point about the overlapping textuality in your actualization of this trailer text. “Anterior text” or “pre-text” are iffy because chronological distinctions are being effaced in both consumption and production. “Primary” perhaps? You could argue that primary and secondary are reversed here for you as a viewer. Let’s find some terms that help us to distinguish levels of textual interaction, rather than the increasingly amorphous word “text.”

Jonathan Gray's picture

work v. text

Ellen, thanks. But I think Barthes’ work-text distinction offers us the answer. Sure, a student who has only read a review hasn’t read the *work*, but for them the review is the *text*. Part of this stems, though, from my own interest in analyzing textuality, which is to understand what work texts do in society. Thus, my interest is in what version(s) of the text are picked up by popular culture and audiences, an interest which ultimately doesn’t mind too much whether the audience is reading it “wrong”, “incorrectly,” and/or insufficiently. Take Twilight, for instance, or The Hobbit, both of which mean a variety of things to popular culture, and serve certain roles and purposes, many of which stem from trailers and hype by large number of “audiences” who aren’t watching the films or reading the books, nor will they ever do so. It could be easy to write those readings off as “wrong,” but if thousands or millions of people are creating meanings about those films based on those paratexts, then the text for a great many people is entirely paratextually-generated. The work itself will often enjoy special treatment and a special place at the top of a value hierarchy, but only for some viewers.


No one has reviewed this post… but you need to login to submit feedback