The notion of the "slingshot ending", which John Clute has appropriated as a critical term of art (see here), began in a thematically-appropriate small way as a remark by Kim Stanley Robinson in a blurb on a Wolfe novel (I believe Exodus from the Long Sun), about how a Wolfe ending seems, right at the end, to open up vast new vistas, to slingshot the readers out into new possibilities of story which are, however, unmapped.
The Gene Wolfe Literary Podcast's marvelous series on The Fifth Head of Cerberus had a slingshot ending.
After carefully building a stunningly original reading — Glenn has called it conservative, but in some ways it is really quite radical when compared to the consensus readings — that the Annese never existed, that they are only a "new planet myth", used to explain and justify various things about the decaying, despotic and violent societies on the twin planets — you have, in your final episode, a radically different interpretation on, one which is only fleetingly compared with the GWLP consensus reading (which, of course, contains disagreements between its two major developers, and some admitted areas of uncertainty); and then we are left.
And Glenn said he can't imagine a fitter ending!?! Say, rather, it was a fitting middle: a prelude to rereading the novel, trying to sort out these two, utterly different, both brilliantly persuasive readings. (I am not, repeat not, seriously suggesting this; but I would love to hear it.)
All of which is to say that the narrative arc of this season of the podcast is structured like a Wolfe story. Like in a Wolfe story, we have many clues, but no final answers. The seeming complex interpretation is undermined, right at the end, but something which throws all we have learned into doubt. And the reader is left to do the work to make sense of it all.
I would like it to be like a Wolfe story in one further respect: that everything is true, that somehow we can create a reading in which everything claimed — in this case, both the GWLP consensus reading and Marc Aramini's reading — is all somehow true at once — all literally true, as Marc is fond of putting it. But unlike in a Wolfe story, I don't think this will be possible.
So what, then, are we to make of these two readings?
The first thing that strikes me is how different they are, not just in conclusion (although surely that too) but in methods. They start from such different places that it is hard to imagine how to even judge between them. If you start by looking for thematic unity, and focusing on evidence in a historical way (possibly my favorite bit of the entire podcast series was Glenn's going through the evidence for the Annese's existence with his historians' hat on), then the GWLP consensus reading will emerge. If you start by looking for symbolic unity and focusing on the way that parts reflect the whole, then Marc's reading will emerge. Entire categories of evidence will be treated differently: Marc takes "A Story" fairly straight, even while seeing it as about John V. Marsch too; GWLP takes "A Story" as the fevered product of a damaged mind, without, ultimately, any real historical evidentiary value at all.
How do we compare two readings, one of which starts with an assumption of universal literalness and symbolic shadows cast by the little upon the great (have they but courage equal to desire), and another which starts with an assumption that many of the characters are lying, that whole chunks of the text are not symbols but errors or delusions or self-comforting myths? How do we stack up Marc's subtle mis-en-scene's with Glenn's brilliant historian's catalog for the barely-existant myth of abos?
Both readings had very strong interpretations of parts of the text — Marc is, for instance, particularly convincing on the recurring symbolism of legs and the parallels to trees and butterflies; I found the GWLP consensus reading of Liev's postulate (at least what I took it to be from earlier: the way it was described in episode 71 was different than my memory of it, which was that it was the literal opposite of Viel's hypothesis: for Veil, the humans are actually abos pretending to be human; for Liev, the abos are just humans whom others pretend are abos) far more persuasive.
I am tempted to say that perhaps those who read the story as ultimately ambiguous were right after all. Perhaps, we might say, they had some presentiment of this ultimate predicament.
How else might we decide?
One way is to step outside. Marc, in private email correspondence, forwarded me excerpts from two Wolfe interviews (different from the one read and analyzed on-air in the podcast) that seem to weight in pretty strongly on his side — particularly the one where Wolfe asserts straightforwardly that Marsch is replaced by a Shadow Child. — But, of course, Marc also correctly notes that Wolfe interviews are not to be trusted. And ideally (as Marc would be the first to point out) a reading should be fully provable from the text alone.
Another way, however, is to look at what emerges. Which reading produces a better novel?
And here I must admit I am tending towards the GWLP consensus. To explain why will require a digression or two.
A great writer always has certain tendencies they must avoid — temptations to which they are prone, threats they must escape — which are shaped by the nature of their genius. Charles Dickens, for instance — to take a writer whose place in the pantheon of great English novelists is fairly undisputed, as well as a personal favorite of Wolfe's — continually threatens to topple over into sentimentality, into melodrama, and into propaganda. These don't make him less than a great writer: the way he is great involves dancing on the edge of those faults. But he does, at times, topple in. They are the failure modes to which his genius is prone. Similarly, perhaps, Proust's failure modes are purple prose and tediousness: but, again, those are the marks of his genius, not refutations of it.
Wolfe's major failure mode is the writing of mere puzzles. (It's not his only failure mode. I suspect his other failure mode, one he approaches less often but to which his genius shows a basic temptation, is that of the catholic propagandist.) I think there is something greater about a novel than a crossword puzzle: partly its expandability, its inexhaustibility — and it's applicability, to use Tolkien's term. Crossword puzzles, at some point, are solved. A good novel has more depths than that. But Wolfe's temptation — and, like the above examples, this is not a refutation of his genius, but rather the way in which his genius manifests itself — is to make his stories and novels sovleable. The threat is tipping over into something can be simply solved in the way that a puzzle is solved. In the way that no solutions to puzzles will ever finish off David Copperfield or In Search of Lost Time.
For all that some details are better read by GWLP than by Marc, I think that, on the whole, Marc is better at making sense of all the little clues and details, his hints and shadows.* I think this symbolic readings probably fit best with Wolfe's spirit. My guess is that it is along the way to the reading that Wolfe ultimately intended.** (I don't think he's all the way there yet, as evidenced by the fact that there are still issues about which he is fuzzy: the best reading will, presumably, have more solid answers for (e.g.) what Liev's post-postulate is.) But I think he's on the right track.***
Glenn and Brandon's reading, in contrast, does not explain all the clues. But it leaves us, I think, with a richer novel. (Themes, ultimately, are richer than puzzles.) The Fifth Head of Cerberus, in their reading, is a novel not about the lifecycle of an (ultimately imaginary) alien, but about the violence humans do to each other and the excuses they invent (and believe) to enable them to do it.
Perhaps it is simply that I want the GWLP consensus reading to be true. It seems to make it a more interesting story. The lifecycle of imaginary aliens is, in the end, not as interesting as the tendency of humans to cruelty and self-delusion and myth.
Now presumably Marc would say that, since the abos copy our tendency to cruelty and self-delusion and myth, that his reading does not exclude this; but it makes it, I think, a pale and less interesting shadow. (The perfect metaphor for this exists, but it is itself more complex than this situation. Nevertheless, for any readers of 80s X-Men comics out there: Marc's reading is like the later retcon about Phoenix copying Jean Grey, and the writer's insistence that this retcon loses nothing; and it is as unconvincing as that defense of that retcon, too.)
Now there is one aspect that I think Marc makes richer than the GWLP consensus: the second novella. For all their marvelous attention to the details of "A Story" in their close reading of it, it mostly drops out of their final reading of the novel, which focuses on "Fifth Head" and VRT. The second novella, in their reading, threatens to become a mere settling of scores and patching together of myths.
But the novel I am rereading (right now) is the GWLP version: because it feels like a better book.
Again: both readings are incredibly powerful. I don't think that any reading of The Fifth Head of Cerberus is going to be persuasive without fully confronting both of them, on their own home grounds of the approach and natures of evidence that makes them persuasive. And I do hold out a hope that there is some way to preserve the best of each these two readings, even if I don't, now, have it. Ironically, if there is such a reading, I suspect that it will coalesce on the "standard" reading to which both Marc & the GWLP take exception: that there are some abos still around, but that Veil's hypothesis is not true; that Marsch is replaced neither by a shadow child version of himself nor by a human VRT, but simply by an abo. Perhaps, in some uniquely triparte Hegelian way, both Marc and the GWLP consensus are antithesis to the thesis of the standard reading, which can all merge into some synthesis.
But that they have left as an exercise for the readers; and I, myself, have no solution to it. If one exists at all.
* Marc's reading does, however, tend to omit, dismiss or explain away the larger, more obvious clues. Often Marc does this by appealing to a proposed two-level reading, in which a mystery contains a (more) obvious as well a hidden solution. But this, of course, threatens to simply ignore the most basic and powerful of the clues.
** One tendency I routinely push back against is the tendency of Wofle interpreters to act as if he is infallible: as if all his clues always add up the way he intends them to, as if he never goofs and weaves in a hint that he does not mean to point the way it actually points. Along these lines, my guess is that something like Marc's interpretation is what Wolfe had in mind, but that he — as an error — did not foresee, and thus did not sufficiently block, the GWLP consensus reading: that the latter is a more-or-less plausible reading of the text, even though Wolfe didn't intend it to be. Because he's not perfect; and, as modern literary theory holds, authors are not in complete control of the meaning of the texts they create.
*** Although it's worth noting that symbolism is ultimately very fuzzy: and that historical evidentiary practices of the sort Glenn brings to bear on the work are specifically designed to guard against the primary failure mode of symbolism, namely, the tendency of humans to see castles in clouds.