Poison River and the vertiginous ellipsis

by aussi disponible en français

In comics,[1]) the ellipsis — a spatio-temporal jump in the narration which happens between panels — is a given which the reader, quite often, ignores. This is normal : if you were constantly distracted by each ellipsis, reading a comic would quickly become tiresome. Most ellipses in comics are “mild”, made to be ignored, but certain ones stand out. This isn’t necessarily because of an error on the artist’s part.
In the context of this discussion, let’s look at Gilbert Hernandez’s Poison River. What first strikes the reader about this work is its narrative density. It isn’t uncommon for a single page to show as many places, times, and situations as there are panels. Abundant dialogue pushes the temporal flexibility of the panel quite far. In comics it isn’t desirable for a single panel to show a real moment like a snapshot. Instead, a “psychological reconstruction” of what happened is shown. Hernandez’s reconstructions, where some panels represent a few minutes narrative time, are examples of the distortion of temporal flux about which Jean-Luc Coudray has written.[2]

The density of narration, the abundance of situations in a limited space, and the compressed representation of time all participate together to give the story a schematic impression. In other words, Hernandez tells his stories in broad strokes, showing details only when necessary. Among other things, this allows him to age his characters significantly in only a few pages or to show the type of large-scale social or political evolutions that would be difficult to notice were the story told “step by step.” On the other hand, these characteristics seems to prevent a certain degree of fluidity in the story.
What is remarkable is that Hernandez makes copious use of dialogue and conversation, which would, at first glance, seem better suited to a more fluid and “natural” narration. But Hernandez’s conversations, rather than going on at length, are often limited to two or three panels, or even a single panel which summarizes the whole situation.
Even more remarkably, these conversations often involve more than two characters. In extreme cases it’s as if a series of characters were reciting lines in front of us, each offering her opinion on or impression of the situation at hand. The reader is left to confront and synthesize these disparate views, because in the next panel it is already six months later. Hernandez rarely makes use of narrative captions ; the temporal calculation (”Six months later…”) is almost always left for the reader to deduce.
We could describe Hernandez’s narration as “hyper-elliptic”, a rather pompous way to say that it is the reader who must connect certain facts together and that the author limits his task to giving sufficient information for aiding comprehension (without excessive effort) of what happens between the panels. More simply, the ellipsis is obvious.

In this respect, Hernandez’s narration has some affinities with pre-war European comics, which some refuse to call “comics” because they are really “illustrated texts.” These quaint engravings, accompanied by verbal captions but lacking balloons and other canonical graphic codes (motion lines, etc.), seem lacking in some of the efficacy with which the comic reaches its goals. How obvious, some will say.
Yes, but… Harry morgan has shown,[3] there are fewer differences than you thank between the “illustrated text” and “comics” as it is often defined. The presence of a word balloon isn’t sufficient to define comics — what about wordless comics ? Similarly, the redundancy created when a caption more or less repeats what the image shows is a flaw only marginally connected to the narrative technique employed. The captions of Blake and Mortimer are often redundant, yet no one describes that series as “illustrated texts.” In fact, we can imagine an “illustrated text” in which redundancy is excluded, where the illustrations and text were always complimentary (for example, James Thurber’s laconic “The War Between Men and Women”). As these artificial categories disappear, it becomes difficult to distinguish in what way Pellos’ Futuropolis would “only” be an “illustrated text”, while Zig et Puce would be a true “comic” (to give two typical examples).[4])

Yet, when one puts Futuropolis and Zig et Puce side by side, an abiding impression remains : the two works use very different narrative techniques. This impression isn’t as superficial as it seems, on the contrary. In a superficial (we might say, “naive”) way, these two works have a similar air to them. The absence of balloons in one, the economy of narrative captions in the other seem, at first, to only have an accidental character, but, in both cases, the sequential operation is similar.
In fact, it is in the division of time that Zig et Puce seems to innovate. Futuropolis presents schematic images like so many stepping stones across the stories’ path. Whereas, the comics of Saint-Ogan give us an impression of fluidity : the steps are more numerous, the time is divided more finely. Like a cartoon, the rapid sequencing of images causes the reader to stop seeing each as a step. Rather, his eye slides from panel to panel, hardly is the last balloon read then the next panel is rapidly surveyed, analyzed, read, and so on.

This impression of fluidity doesn’t come from the introduction of the word balloon. The author creates it by expanding time in the narration. He breaks his action down into more units. This takes more space but, perhaps paradoxically, produces the impression of fluidity because the reader is almost never forced to ask “what happened between these two panels.”
This issue, which seems insignificant, has its importance. It is what, for a long time and still today, associated the “invention” or “discovery” of the word balloon with a kind of narrative revolution. Without denying the importance of the word balloon, in itself it is only peripheral to this type of fluid narration which could exist without word balloons. We can notice it in Töpffer, who already seems to use this narrative mode. The difference between “illustrated text” and comics is therefore reduced, most often, to a simple effect of mode.
Inversely, in Hernandez’s work (and it is remarkable here), we have the word balloons and a remarkable economy of narrative caption, but the rhythm often resembles Futuropolis more than Zig et Puce. Test it out yourself : imagine Hernandez’s panels without word balloons accompanied instead by an equivalent caption. The rhythm is unchanged.

What Hernandez does is not simple. It’s a matter of compressing, compartmentalizing even, the story while giving it the “natural” aspect of a fluid narration.[5] His narrative rhythm is slow but staccato ; he brilliantly shifts between several speeds in the service of the reader’s pleasure.
It is rare to read a comic where the ellipsis is as systematically vertiginous as in Poison River,[6] one particular aspect of this story, among others, which helps make it a major comics work of uncommon inspiration.


  1. Throughout, I use “comics” in place of Turgeon’s “bande dessinée” as he is discussing comics as a formal entity, not any particular cultural line. (Trans. note
  2. Jean-Luc Coudray, “Du flux temporel en bande dessinée” [The Temporal Flux in Comics], L’éprouvette #1, L’Association p.67.
  3. At many places in Principes des littératures dessinées [Principles of Drawn Literature], Harry Morgan, Editions de l’An 2. We could even say that this resemblance between certain illustrated stories and what one generally calls comics is the principal subject of the first book ; this is why Morgan introduces the more general term of littératures dessinées [drawn literature].
  4. For English readers, who may be unfamiliar with the two works in question, here’s an example of Futuropolis and you can read some Zig Et Puce here. An equivalent example in English might be Prince Valiant and Little Nemo. (Trans. note
  5. We will note in passing another recent work such as Le cargo du roi singe by Joann Sfar and Tanquerelle, which has the originality of using, in turn, the two narrative rhythms, though using them in a classic manner : the schematic passages use narrative captions, the rest of the story, balloons. The episode where Professor Bell studies the giant monkey is of the “illustrated text” mode. As for Edgar P. Jacobs, he is a separate case : multiplying the captions but clearly searching for a fluidity between the panels.
  6. We could equally note certain passably acrobatic ellipses in Chris Ware’s work, an author as at ease with very fine breakdowns as with excessive schematization. In this sense, if one allows us this digressions, we could consider the work of Ware as “bande dessinée totale” [complete comics]. We will note in contrast that the ellipses of Ware are generally more regular, therefore easier to read, than those of Hernandez.
Dossier de in November 2006