I don’t believe in right and wrong.

That may sound strange, from a Batman fan; especially strange in the context of The Killing Joke, Alan Moore and Brian Bolland’s prestige-format, psychological and physical showdown between Batman and his arch-nemesis Joker.

Even if you propose, as I do in my recent monograph Hunting the Dark Knight, that Batman and Joker occupy different points on the same spectrum of insanity, rationality, morality and theatricality, rather than operating as binary opposites – that their never-ending battle is based not on their differences but on the fact that they have too much in common– a Batman fan, a Batman scholar, must surely cling to some idea of ethics, reason and truth.

Sure, I do: I think the ethical, moral and reasonable stance is sometimes to accept that there is no absolute truth or ‘right answer’.

The multiple theories and interpretations about what The Killing Joke is about, what it means – even, on the simplest level, what it actually shows us in its more ambiguous panels – add to its richness. It’s not every comic book that can keep us debating, twenty-five years after its first publication, and poring over individual frames using technology that didn’t even exist in 1988. So I’m grateful to people like Grant Morrison and Julian Darius for reawakening my interest in this book and returning me to its pages with their recent re-readings of the story and its conclusion.

I don’t think either of them is definitely wrong. I just don’t think either of them is absolutely right.

To reiterate, Darius and Morrison agree that one crucial panel of The Killing Joke shows Batman killing Joker. They disagree on the method, and sometimes they argue more forcefully that it definitely happened, rather than that it could have happened, but that’s the crux of their argument.

I have no objection at all to the proposal that the ending of Killing Joke is ambiguous: as noted above, I welcome it. Without that provocative argument, I would have left the book on my shelf, rather than studied it again this week. And on principle, I can’t disagree with the idea that all texts are inherently open to various interpretations.

I believe in meaning as a democracy of interpretations. We all have the right to argue for our own reading, in an arena of dialogue. Morrison and Darius have said their pieces – Morrison in an interview, Darius both in a video and in his book, And the Universe So Big. I thank them, genuinely, for their readings. They are not ‘wrong’, and they add to our complex, broad understanding of what The Killing Joke is, what it means: that meaning is created not just by Alan Moore and Brian Bolland as authors in the mid-to-late 80s, but by us, in the 25 years since publication.

But in a democracy of interpretations, I have the right to argue against them.

Morrison stresses that the ending is ambiguous. However, he also strongly implies that it’s deliberately ambiguous: ‘It has to be… that’s what makes it great.’ As such, he is saying Moore intended us to read the conclusion as hinting that Batman killed Joker.

That’s the first point where I disagree with him.

We should note that there are enough slips and stumbles in Morrison’s conversation with Smith to warn us against taking this interview as gospel: both Morrison and Smith, tellingly, flip the names of Batman and Joker during their discussion. Morrison describes the final page as showing that ‘the Joker reaches out’ and Smith agrees that Moore wrote ‘the last Batman story’ – neither of which, obviously, are correct. ‘Of course he did,’ Morrison eagerly agrees:

That was the reason he [was] hired, to write, you know, he did it, and he did it in such a way that, no, it doesn’t…only if you notice, and it’s beautiful.

Morrison is a great writer, and it would be unfair to pick on him for not always managing full sentences or maintaining a clear argument in a radio conversation – my own interview transcripts would look at least as incoherent. But his point is never made clearly, and he is arguing from memory: sometimes false memory (he recalls, for instance, that the book was published almost 30 years ago) and sometimes partial information or plain inaccuracy (he claims he’s the first to notice the ambiguity, but at least one online community was discussing it in 2006).

Morrison suggests that Moore was ‘hired’ to write ‘the last Batman story’; the implication is that DC editors wanted a final word on the Batman/Joker relationship, and Moore deliberately gave them a conclusion that could be read either way.

But nothing seems to support this interpretation. Moore was not, according to the Killing Joke introduction and afterword, ‘hired’ to write the book in this sense: although even in print, there remains another fascinating ambiguity. Tim Sale initially states that Moore asked Bolland what he wanted to do next, and the artist replied ‘the Joker, please’; Bolland, in the final pages of the book, contradicts him. ‘There’s a minor detail that Tim got wrong…it was me that asked Alan to write the book and not the other way round.’

In another interview, Bolland develops this story further, claiming that in 1984, Dick Giordano offered him carte blanche to work on any DC property, and he requested a Batman/Joker prestige book with Moore.

It’s a minor, but again, a significant detail, in that the origins of the book itself, like the origins of the Joker in the book itself, are never certain, never confirmed.

Something like that happened to ME, you know. I… I’m not exactly SURE what it was. Sometimes, I remember it one way, sometimes ANOTHER… if I’m going to have a PAST, I prefer it to be MULTIPLE CHOICE!

So the real-world history of The Killing Joke slips and resists, never quite fitting into canon and continuity, and that deliberate ambiguity around the book itself is a factor I’ll return to in my conclusion.

However, nothing in either the introduction or afterword suggests that Moore was ‘hired’ to write the script in the way that Morrison implies: that he was given a commission to provide the ultimate Batman/Joker story, and he chose to deliberately subvert or complicate it by making the ending ambiguous.

There are a number of contextual aspects that can be marshalled to support Morrison’s argument, and it would be only fair to mention them here. Moore did, at the time of The Killing Joke (if we locate its writing and production sometime between 1984 and 1988) play with ambiguous images and narrative devices. Watchmen, most obviously, leaves the ending ‘entirely in [our] hands’, while ‘Whatever Happened to the Man of Tomorrow’ ends with a wink to the reader.

Additionally, Moore was clearly keen, in the mid-1980s, on the idea of providing final endings for iconic characters. In his introduction to Miller’s Dark Knight Returns, he praises the book for ‘introducing that element without which all true legends are incomplete… that element is time.’

All of our best and oldest legends recognise that time passes and that people grow old and die…with Dark Knight, time has come to the Batman and the capstone that makes legends what they are has finally been fitted.

‘Whatever Happened to the Man of Tomorrow’ opens in the then-future of 1997 – again, introducing that element of time – and provides us with a final story for Superman. However, the wink at its conclusion opens up the possibilities of a postscript, just as Watchmen concludes with the possibility that Rorschach’s Journal could blow the whole thing open, and indeed, just as Dark Knight Returns, which Moore so admired, ends with another Superman wink, and the revelation that Batman’s death was faked.

So Moore was, at the time of writing The Killing Joke, clearly interested both in playful ambiguity and in final endings, which lends immediate credence to Morrison’s theory from an auteurist perspective. It would, we could argue, be ‘typically Moore’, or typically mid-80s Moore, to both close and open the ‘final story’ on Joker and Batman.

However, if we want to argue that Moore intended the reading Morrison identifies in The Killing Joke, one key piece of evidence weighs in the other direction: Moore’s instructions to Bolland, in the form of the original script.

Page 46, frame 3:

He and the Joker are going to kill each other one day. It’s preordained. They may as well enjoy this one rare moment of contact while it lasts.

Frame 5:

They are now both helpless with laughter and have collapsed forward onto each other, both ragged and bloody, both holding the other up as they stand there clinging together in the rain.

Does the author’s intention decide the meaning of a story or a scene? I don’t believe so; I see it as one more valid interpretation among many.

However, if Moore intended the scene to depict Batman enjoying one last, rare moment of connection with Joker, and specified that they are clinging to each other in an embrace, what we can’t argue is that Moore also intended to show Batman breaking Joker’s neck, or that he meant the image to be ambiguous: unless we want to suggest that either he was hiding the secret meaning from editors, and from Bolland, and perhaps from himself. Again, that’s a possibility I’ll return to in my conclusion.

Is it conceivable that Moore prepared a ‘safe’ version of the script? It’s never been officially reproduced, to the best of my knowledge, and the only extant versions are badly-scanned, uploaded bootlegs, so it seems unlikely that this document is some kind of ‘cover’ for his true intentions. (Dave Gibbons confirms, in his own introduction to Across the Universe, Moore’s DC stories, that ‘the reader never sees’ Moore’s scripts).

Equally, it seems like conspiracy theory to assume that Moore would write this script – which includes multiple, chatty asides to Bolland, and is clearly meant as a set of directions and suggestions to a friend or close colleague, not a public document – and then telephone the artist to let him in on the real story, asking him to make it clear that Batman could be throttling Joker on that last, crucial page.

What, then, of Bolland’s enigmatic afterword? As Darius notices , the artist remarks ‘People seemed to find the last page of the story ambiguous, so before I conclude this text, remind me to reveal what actually happened.’

In his last paragraph, Bolland announces that he’s going to reveal the truth:

…as our protagonists stood there in the rain laughing at the final joke, the police lights reflecting in the pools of filthy water underfoot, the Batman’s hand reached out and…

It’s a deliberate tease: Bolland has opened his afterword with a warning that ‘if I should stop in mid-sentence it’s because I’ve run out of space.’

Morrison also picks up on this as confirmation that the ending is intended to be ambiguous, with Kevin Smith agreeing that Bolland would know.

But what does Bolland actually know? He knows, by 2008, when this edition was published, that ‘people seemed to find the last page of the story ambiguous’. We’ve already seen that this is the case, as fans were discussing it online in 2006, and quite possibly earlier. Bolland admits it’s ‘additionally cool to me that all the writing in this book has been given over not to writers but artists [himself and Sale], a breed of people not known for their ability to string a sentence together.’

The reason the artist takes centre stage here is, of course, the same reason that the Titan Books volume Watching the Watchmen is authored by Dave Gibbons, not by Alan Moore, and that John Higgins is the only member of the original team involved with Before Watchmen. Alan Moore would not write an introduction to a DC Comics reprint of his own work.

Bolland is right that writers usually get the attention. My feeling is that, understandably, he quite enjoys this platform – he mentions the thrill of being introduced by Tim Sale, just after watching Heroes with his 11-year-old son – and this rare opportunity to entertain the reader. But he’s an artist, not a wordsmith. He’s no Alan Moore – he’s not going to write a show-stopping essay like the introduction to Dark Knight Returns –  so what can he do with his moment in the limelight? He can perform a classic theatrical tease.

What Bolland knows, I suspect, is not that Moore intended the scene to be ambiguous, but that fans have read that scene as ambiguous. So he decides to play with that knowledge. I believe his afterword is a comment on and knowing manipulation of the fan response, rather than a revelation about the underlying ‘truth’ of the scene, or even about Moore’s original intention.

A further problem with Bolland’s hint – and with Morrison’s use of this hint as evidence of what ‘really happened’ (based on Bolland’s creative authority; ‘he should know’, as artist) – is that the finished panel itself only shakily  supports the interpretation that Batman, in Morrison’s words, ‘reaches out and breaks his neck’.

We’ve seen that the script offers an entirely different reading, but even Bolland’s art, his reading and visualisation of Moore’s script – and even in its cleaned up, re-coloured and partially-redrawn 2008 incarnation – clearly shows Batman’s right hand making contact not with Joker’s neck, which is lit white in the approaching headlamps, and exposed above his collar, but with his shoulder or upper chest. Again, if Bolland had wanted to suggest that Batman’s hand was anywhere near Joker’s throat, he had the opportunity to tweak and clarify –‘every page’, he tells us in the afterword, ‘has something slightly different on it from The Killing Joke of 20 years ago…’ but, despite his teasing at ‘what really happened’, he chose not to confirm it in the new art.

Or at least, he chose not to confirm specifically that Batman breaks Joker’s neck. But does that rule out the idea that Batman kills Joker, in some other way? It makes Morrison’s theory harder to sustain, but Julian Darius’ And the Universe So Big submits and argues for another, more complex interpretation, with the same result.

The key difference between Darius’ theory and Morrison’s is the method of the killing. Rather than argue that Batman’s hand on Joker’s chest signifies ‘breaking his neck’, Darius advances the explanation that Batman is thrusting Joker’s poison-tipped needle – the one he used on the fairground owner – into his enemy’s skin, where it paralyses him instantly.

It’s more plausible in terms of that panel and the position of the two silhouettes, but by consequence, Darius has set himself the task of explaining how Batman got hold of Joker’s weapon. To his credit, he does this with diligence and insight. I don’t agree with him, but I admire his argument and the way he marshals his evidence.

To prove his poison-needle theory, Darius has to go back to another ‘ambiguous panel’, a few scenes earlier. This one is page 42, panel 3 in the script – but, with fitting ambiguity, it becomes panel 4 in the finished comic, as Bolland adds an extra frame to show the Joker’s concealed knife flicking from up his sleeve. Bear that inconsistency in mind, as it has important consequences for what follows.

This is the frame in question, on the finished (2008) page.

‘This is a crucial sequence,’ Darius suggests, ‘because, while readers would be forgiven for not noticing it, the Batman stares at his right hand while recovering.’ In Darius’ view, this is the moment when Batman realises he’s picked up the poisoned needle, which he kicked out of Joker’s hand on the previous page.

The next panel ‘shows Batman reaching back with his left hand to catch the Joker’s wrist. But now his teeth are suddenly gritted angrily, and he’s still staring forward at his open right hand.’ This leads Darius to the conclusion that, at least in theory, Batman may ‘have one of the Joker’s poisoned needles in his hand’ and subsequently ‘kill the Joker’ at the end of the book.

Sensibly, Darius doesn’t push the theory too hard in his book, and admits that ‘none of this is conclusive…there’s no way to prove that he’s got the Joker’s needle device in his right hand, but it’s certainly an odd page when you really look at it’.

However, in his recent video he goes further – I believe too far – and, encouraged by Morrison’s announcement, states boldly that ‘Batman is a murderer, he did kill the Joker at the end.’

‘This,’ Darius confidently agrees, ‘is completely true.’ And here, again, I think he oversteps the mark, enthused by Morrison’s very public confirmation of his own theory. ‘He’s absolutely right…but I think that he’s wrong about the method…he suggests that Batman snaps the Joker’s neck…Batman is using the Joker’s venom from his needle-device in his hand, to kill the Joker.’

Now the ambiguous panel has, to Darius, become unambiguous. Batman kills the Joker – there’s no ‘may’ about it, no ‘none of this is conclusive’, no admission that there’s ‘no way to prove’ the theory. He goes on to reiterate his explanation from the book, with less subtlety and more passion:

He kicks the Joker’s needle out of his hand…and he stops and he stares at his hand. And it’s like, half the page in this fight scene, he’s staring at his hand. Why is Batman staring at his hand, in the middle of this fight scene, for half a page?

Just previous, Batman had kicked that needle out of the Joker’s hand…so, is it possible that he got that needle caught in his glove, and he’s staring at it? Can you imagine, in the chaos of the fight, you lift up your hand, and there’s a fuckin’ needle there?

After that conflict with the Joker, where Batman maybe picks up his needle, suddenly you don’t see his hands, until he slaps the Joker on the back, and suddenly the Joker acts as if he’s just been hit with his own Joker toxin.

Darius is absolutely right about one thing. This is an(other) odd page, and it can be interpreted in numerous ways. His theory is an interesting one, backed up with close references to the text and a persuasive argument about the symbolic motifs that might clue us in to this conclusion: the raindrops resemble the needle device, the book returns consistently to images of hands, and other ambiguous panels seem to echo the needle motif – not proving the theory, but subtly guiding us towards that interpretation.

And, as I suggested above, it’s important to be fair in a democratic arena of meaning-making, and to provide evidence for the opposing view, if you believe it’s there. Darius’ argument is supported by the fact that Moore, during this period, extensively used visual echoes and symbolism. Watchmen has many panels that focus on hands – not least clock hands –  until the payoff in the final panel, that the key to busting open the whole conspiracy is ‘entirely in your hands’.

So these subtle visual resonances wouldn’t be remotely out of keeping with Moore’s style at the time, and this reading is solidly supported by an auteurist approach.

In his energetic video response, Darius is inevitably less careful than in his book. It’s not strictly accurate to state that Batman is looking at his hand for ‘half’ the fight scene, for instance – it’s two panels out of seven – and we do see Batman’s hands again before the conclusion, once when he balls his right fist on the same page, and once when he holds open his left palm while offering Joker a final chance. (Darius comments on this image himself, in the book). Of course, it’s also incorrect, or at least very implausible, to suggest that Batman slaps Joker ‘on the back’.

Equally, I believe Darius is mistaken to suggest that Batman, ‘no longer in rage, but calm, delivers this blow to the Joker, which causes this rigor mortis effect.’ He’s clearly laughing along with Joker.

Again, a critic can be excused for minor errors of interpretation and expression in improvised responses. But even in its more carefully-argued form, Darius’ theory presents its own problems, which he can’t answer. For Batman to hold the needle device in his palm in one panel, and slam his fist into Joker’s stomach in the next, he would have to somehow move it from his palm, or form a fist without allowing the needle to pierce his own skin. In the next panel, we see his fist in close-up, gripped tightly. Where’s the needle, in this context? To explain that Batman has already secreted it elsewhere between panels is to fall back on the ‘he’s just Batman’ theory that can explain any plot hole or inconsistency.

In terms of what we know of the character – even if we disregard the valid argument that Batman would follow the instructions of his old friend Gordon to bring in Joker ‘by the book’ – do we really have to imagine that Batman needs a weapon he’s picked up off the ground? We’ve already seen that he has a vial of alkaline solution to calm Joker’s acid spray, accessible within the space of a single panel.

Whatever your definition and understanding of Batman, it surely includes the idea that he’s an accomplished scientist, armed with technological inventions, and that he wears a Bat-belt loaded with a range of weapons:  there’s a certain symmetry in using Joker’s poison against him, but it seems an unnecessary step.

Similarly, do we really have to explain where the poison-device has disappeared to, as Darius does, with such an extensive theory (‘could it have arced high enough to hit the wall and come to rest along that wooden railing?’) We don’t see what happens to Joker’s microphone on the previous page – it simply vanishes between frames 1 and 2 – and we don’t feel the need to come up with paragraphs of back-story to track its possible trajectory. This is a lot of rationale to justify one small, puzzling detail.

So, what’s the explanation for that second odd panel? Well, according to Moore, there’s nothing especially strange going on.

Page 42, Panel 2 [panel 3, in the finished comic]

The Batman… is just getting his mask straight with one hand… we see the Joker towering behind him with the knife in his right hand.

Panel 3

Same shot, same angle. Without looking round, The Batman’s left hand shoots out back, expertly catching the Joker’s knife-bearing right arm as it descends towards him…

The key here, I believe, lies in the fact that The Killing Joke is a collaboration. Comparing the script and the finished page reveals that Bolland frequently omits and ignores Moore’s suggestions, especially during this sequence: Moore requests grotesque  details, nursery-rhyme decoration and Bill Finger style props in the background, but Bolland opts for almost blank backgrounds, with only the simplest suggestion of shadows, frames and stripes. Ironically, we might expect Morrison to remember the process of rough translation between comic book writer and artist, rather than assume that everything we see on the page is Moore’s intention: Morrison’s own, very similar instructions that the Mad Hatter’s chambers in Arkham Asylum should be filled with crowded, busy décor and the ‘giant props’ of traditional Batman stories were ignored by Dave McKean.

McKean and Morrison clearly experienced tensions during the making of Arkham Asylum, and Bolland and Moore also had their differences. Bolland states in the afterword that ‘the script for The Killing Joke was very good, but I must admit I had to grit my teeth a couple of times during the drawing.’ Again, remember that, because in my view, Bolland holds the explanation to this entire mystery.

Darius generously entered into conversation with me on Twitter during late August 2013, explaining his own theory further. Inevitably, the 140-character format limited his replies, and my questions, but he made one point repeatedly.

If you look at that page, he’s got his other hand there. Both panels are inexplicable otherwise.

He’s got his other hand up to his face / first hand. Reading it, it’s very odd.

My point was, both of Batman’s hands are in play there. Not just ‘it must be in the punching hand’.

It’s a valiant attempt to explain a confusing couple of panels. But the truth is, I believe, that there’s a far simpler explanation. Darius draws our attention to the fact that Bolland’s afterword foregrounds the ‘ambiguous panel’ of Batman slapping Joker in silhouette, but he ignores the sketches at the end of the book.

The single remaining prelim from The Killing Joke – the only preparatory art that hasn’t been sold to collectors – is, conveniently, page 42, where Joker attempts to stab Batman in the back.

And panel three, with Batman adjusting his mask, which Joker has pulled over his eyes a few moments ago, shows that Batman only has one hand to his face. When he first sketched the page, Bolland was following Moore’s instruction, that ‘The Batman… is just getting his mask straight with one hand.’

So Darius’ puzzlement over why Batman has two hands to his face is based on a change between the prelims and the final art. Why, then, did Bolland make that alteration, deciding to show Batman’s left hand adjusting the mask, and his right hand held up? I believe the answer to that can be found in the script, and the fact that Bolland’s job was to interpret what Moore suggested, to translate from text to visuals, in some cases making his own creative decisions for the sake of clarity.

Panel 3

Same shot, same angle. Without looking round, The Batman’s left hand shoots out back, expertly catching the Joker’s knife-bearing right arm as it descends towards him…

Just as he chose not to clutter the background with distracting props and décor, I believe Bolland realised that Batman’s hand shooting out, ‘without looking round’, after he was hit on the head with a 2 x 4 and temporarily blinded (he’s groaning and bent double, clearly injured), needed further explanation. What is Batman palming, then, as he adjusts his mask, and staring into as he snaps his hand back precisely to catch Joker’s killing blow?

A mirror. A hand mirror.

Does this explanation work perfectly? No. We don’t see any previous or subsequent evidence that he’s palming a mirror – though if we wanted to, we could argue that the consistent motif of ‘reflection’ throughout the book signals exactly what Batman is holding in his hand. It’s no more implausible than Darius’ theory that the raindrops are intended to clue us into the use of the poison needle. We could even point out that this scene takes place specifically in a Hall of Mirrors, and that the previous pages have shown Joker’s distorted reflections, and Batman leaping through a pane of glass. Thematically, the underlying message of the story is precisely that Batman and Joker are reflections of each other. If we wanted to go in that direction, there’s contextual evidence enough to support it.

But to me, it makes more sense that Bolland simply decided Batman, the man with a thousand pocket devices, was palming a mirror, in order to explain how he could know the precise location of his enemy’s descending blow, and intercept it without looking back. The prelim art shows Batman adjusting his mask with his right hand, and the final art shows him pulling it down with his left, so that Batman can already have the mirror in the right palm. It’s a simple decision by the artist, intended to smooth over the continuity between panels.

In fact, the evidence is right there on the first page of the script, where Moore explicitly advises that Bolland should go his own way if he feels it works better, and shouldn’t feel constrained by the text.

As with all my visual suggestions… please don’t feel bound in by them in any way. They’re only meant as workable suggestions, so if you can see a better set of pictures than I can…then please feel free to throw out what I’ve come up with and substitute whatever you feel like.

I think that’s exactly what happened here. It’s not complex or profound, but I think it’s more convincing than any alternative argument about how the poison needle arced high, hit the wall, rested on a wooden railing, then was somehow clenched in Batman’s hand, without piercing his skin, while he drove that same fist into Joker’s gut.

And what of the other key evidence in the ‘Batman kills Joker’ theory, cited by both Morrison and Darius, that we know Joker is dead because the laughter stops and the light goes out?

Well, it’s surely inevitable that the light must go out. Thematically, it signifies the end of the temporary connection between Batman and Joker, the severing of their bond and the fact that, as Joker predicts in his final speech, he won’t ever make the leap of faith between insanity (the asylum) and rehabilitation (the city beyond).

On a simpler level of design and visual narrative, the light must go out to complete the circle between the first panel and the last, and to lead into the identical front-and-end pages of puddles (without headlights – they appear faintly coloured at the start of both the 1988 and 2008 editions, but Moore’s script specifies that this initial frame is dark, and that the lights only appear in frame 2).

On the most obvious, literal level, the headlights are extinguished because the police car has either turned them off, or left the location. There are enigmas in The Killing Joke, but I don’t count this among them.

As for the laughter ending, Darius himself unknowingly provides the explanation when he discusses the earlier sequence of Joker’s cabaret song.

What’s so brilliant about this is that it takes advantage of the unique principles of the medium of comics. In comics, the reader controls time, which generally occurs between panels. Two panels might show a man walking down the street, but it’s usually not clear how much time this takes – it’s up to the reader to imagine.

On an earlier page, we’ve seen Batman interrogating the underworld, from petty thugs to the Penguin. He’s terrorising a mugger in one panel, and after two intervening frames, he’s in Penguin’s cell. Two panels on, he’s in a red light district, and one panel after that, Batman is perched on a rooftop.

Do we assume that each panel must take place one second after the last – especially as they’re punctuated with the ‘Bdump’ of a carnival ride banging through doors – and that Batman must be moving with supernatural speed?

No. We assume this is a montage that compresses a longer stretch of time into a short space.

Why, then, must we assume that the final panels of the last page are dictated by a strict rhythm of passing moments, and that Joker’s laugh ends abruptly? We know, incidentally, that this is a shared laugh, and that the massive ‘HA HA HA’ filling the air is an explosion of strange mirth from both Batman and Joker – so when the sound effect stops, they have both fallen silent. Has Batman also been paralysed or had his neck broken? If not, why does his laughter stop?

We can also see that the siren of the approaching police car finishes with an equally sharp cut. One panel it’s filling the space, the next it’s gone. A siren does not suddenly fall silent, any more than raucously shared laughter. Has the car vanished, have the drivers died at the wheel?

No. Time has passed. More than a moment has passed.

For all we know, the final panel, of rain hitting a black surface, with the light extinguished, could be some time later, after the GCPD have cleared up the scene and departed. It could be hours afterwards, late into the night. It could be a different location of the fairground – we can’t see any other visual cues to tell us that we’re looking at the exact same spot. It’s just rain, falling all night, as it did in the opening frames; a sense that, contrary to the theories of The Killing Joke as a final Batman story, or a depiction of Joker’s murder, everything will go on as before.

And that resolution-without-end is surely at least as disturbing as the idea of Batman as murderer. Everything will go on: the whole rotten circle, in repeat. In this exceptional episode, Batman offered a one-time chance at redemption and rehabilitation, and Joker demonstrated that he couldn’t, wouldn’t take it. So the cycle will continue. Batman will put Joker in jail, ‘by the book’, and as ever, Joker will escape. Once again, he will kill, harm and traumatize – in fact, he’ll go on to kill Jason Todd very shortly after this story, in continuity terms – and once again, Batman will put him in jail, knowing he’ll be out again, to kill again.

The cycle will continue, a vicious circle like the poison needle, like the concentric rings of rain, like the headlights of the Batmobile and the police car, like the moon above the harbor. Do we need more than that, in terms of grim resonance? The ultimate Batman/Joker story is precisely that it never ends, that the circus carries on, that these two clowns in costume, these funhouse reflections of each other, will continue their endless dance as others die around them.

That’s the dynamic I discuss at length in the penultimate chapter of my 2012 book, Hunting the Dark Knight. For Batman to kill Joker would, on one level, be a far more comforting conclusion than this macabre merry-go-round where Batman plays through the same routine of locking him away, knowing he’ll escape, and using Joker – and his many innocent victims – as a justification for his own continuing ‘mission’. In my reading, Batman needs Joker to define himself against, to prove to himself that there’s someone madder, more twisted, more destructive, more aberrant out there; to prove to himself that the world needs a Batman.

That’s explanation enough, for me.

Well, maybe we need one other explanation. Both Morrison and Darius pick on the title as further evidence for their theories. Why is it called The Killing Joke, if Joker isn’t killed at the end?

Again, the answer is perhaps disappointingly simple. Morrison and Darius will no doubt remember that Moore poaches freely from popular culture: the Watchmen chapters ‘At Midnight All The Agents’, from Dylan, and ‘Absent Friends’ from Elvis Costello, for instance.

Killing Joke are an English post-punk band, formed in 1978. Their debut album in 1980 was called Killing Joke. They’re exactly the kind of outfit Moore would have known of, and perhaps admired – they were even referenced in reviews of the comic, in the late 1980s. There’s as much mileage in asking why The Killing Joke has that name as there is in interrogating the chapter titles of Morrison’s Zenith Phase III (‘News from Nowhere’, ‘Stairway To Heaven’). They’re part of the particularly-1980s, arguably postmodern trend for quotation and cross-referencing from eclectic sources.

It’s more than tempting to read complex explanations into those ambiguous pages: in fact, the 2008 volume encourages it. The entire volume enshrines and celebrates ambiguity – not in the central text itself, whose confusing moments, I’ve argued, can be resolved through simple explanations – but in the paratexts that surround it, in the introduction and, in particular, the afterword.

We’ve seen that Tim Sale’s version of Killing Joke’s origin contradicts Bolland’s. Why would an editor retain that awkward correction, with the second contributor telling us the first is mistaken, unless the aim was to encourage a sense of ‘multiple choice’ and uncertainty around the book and its history? Why reprint the panel of Batman reaching out to Joker above Bolland’s afterword, and in neon green on the front cover as a hidden image under the dust-jacket, unless to draw our attention to it as a key image, loaded with significance? Why not do that in the original, 1988 print of The Killing Joke, which was equally lavish in its production, just as carefully-designed and intricately-planned, with its embossed cover and clever frontispieces?

Because there was no sense, at the time, that page 46, frame 5, had any particular meaning. Because it wasn’t regarded as an ‘ambiguous panel’ or a key to the secret of the story. It acquired that significance only over time, as fans started to study it and debate it, and gave it a power and complexity that came from readers, not from the creators.

Again, the key figure in this process emerges, perhaps surprisingly, as Bolland. He could have drawn Batman and Joker more clearly, illuminated from the front, to show that they were simply embracing in desperate laughter. He chose to draw them in silhouette.

Those silhouetted figures on page 46, frame 5, became a gift to DC Comics, who were able to remarket the book with a subtle emphasis on that panel in the production design. In turn, they became a gift  to Brian Bolland, who was able to hold centre stage in his afterword with a 20 year-old mystery and keep it to himself, gaining a secret power, a sense of textual authority, often denied the comic book artist (especially when the writer is Alan Moore).

They’re also a gift to readers – including celebrity fans like Grant Morrison and Julian Darius, as well as the less-known and anonymous fans from discussion boards, many years before this story grabbed headlines. For those readers, the silhouetted figures lend the text a new puzzle at the very least, and at best, a whole new level of complexity. Darius spends 17 pages of his book discussing ‘the ambiguous panel’ and its implications. Without that mystery to unpack through his fascinating, close analysis, his discussion would be far thinner, and The Killing Joke would arguably not even give him enough material to justify And The Universe So Big.

And finally, Morrison is absolutely right on one level, though I disagree with many aspects of his argument. I don’t believe the possibility of Joker’s death was intended by Alan Moore; I believe it’s been read into the text over the last twenty years. I have argued that it’s a creation of the reader, not the author, and that the ambiguity of the images is due to Brian Bolland’s creative decisions, rather than to Moore’s script.

As such, Bolland played more of a role in authoring The Killing Joke than we might assume, and the ambiguity is his work – perhaps his deliberate work. We’ve seen that he ignored Moore’s careful instructions on many panels, and (as Moore suggested he could) went his own way. Rather than praise Alan Moore for the complexity and cleverness of his conclusion, more credit should perhaps go to Brian Bolland for putting this twist on Moore’s text, opening up the script to new possibilities through his use of silhouette and the positioning of the two figures and their hands. As such, my earlier reading of Bolland’s theatrical tease in the afterword underestimates his authorial importance: perhaps he really was hinting at an enigma he, personally – whether deliberately or unconsciously – added to the text. I’ve argued that the possible death of the Joker was never Moore’s intention, but we don’t yet know if it was something Bolland explicitly, subtly, wanted to suggest.

Grant Morrison was not the first to identify this potential meaning, and neither was Julian Darius. We can’t know who the first fan was who turned to his or her friend and said, ‘have you noticed, he could be murdering Joker here?’ It could have happened in 1989, long before widespread internet access. But what Morrison says, fundamentally, is true. That ambiguity, whether we agree with the’ Joker’s death’ theory or not, enriches the text. By adding a further layer of complexity, it inevitably makes The Killing Joke a more interesting book. Whether we agree with him or not, Morrison and critics like Darius have encouraged other fan-scholars like me to return to The Killing Joke and study it again closely, arguing for our own readings, and no doubt appreciating it afresh.

Perhaps Moore did intend that ambiguity on some deep, unconscious level, and perhaps Bolland picked up, somehow, on what Moore really meant the last-page panel to depict, and drew it accordingly, including the potential for interpretation.

We can’t be sure, until either author or artist gives us a more explicit, unequivocal history of the book’s development. My own reading of The Killing Joke is much the same as my reading of it in 1988; I have always interpreted the scenes as showing that Batman and Joker share a brief moment of connection, laughing about their shared madness and grimly embracing their similarities. It’s a disturbing and empty ending, in many ways, as I’ve already suggested; not least because, as Barbara recognised in a later story, it seems very much that the two men are laughing about her, or at least laughing despite the fact that she’s been sexually assaulted and traumatically injured. It’s an ending that doesn’t cast Batman in a flattering light.

It’s not a happy ending, but it’s an ending that I’m satisfied with, because I think it reveals something uncomfortable and true about the Batman. He is more Batman than Bruce Wayne (we only see him unmasked for a single page here, in full costume) and he connects with Joker, who barely retains any original personality, far more than he does with Barbara Gordon, who he still sees as a librarian playing at being ‘Batgirl’. He and Joker are both born from tragedy and bad luck. Both have, largely, lost touch with normal life. Both understand what it is to be possessed by a theatrical persona, more than Jim or Barbara will ever know – maybe even more than Selina and Dick, who still remain relatively grounded and sane, will ever know.

That’s how I originally read it – as I was doing my A Levels at the time, perhaps I didn’t see it quite so clearly or in such detail as I do now, having written a PhD and numerous academic studies of Batman – but that, in essence, is how I originally read it, and that is how the final scene still plays to me now: as a grim confirmation that Batman and Joker are both grotesque figures, bound into a dance of death.

Contrary to Batman’s big speech, this relationship won’t ever end in the death of the Dark Knight or the Clown Prince of Crime, as they can never kill each other: they’re like gargoyles above the city, looking down on the little people. But their brutal, never-ending cycle of chase, imprisonment, release and conflict will inevitably trample and maim the supporting cast around them, from the hundreds of innocent civilians poisoned by Joker’s toxin to Barbara Gordon, shot in the spine, and Jason Todd, beaten to a pulp. Batman will visit the victims in hospital and mourn his fallen sidekicks, but ultimately he has more in common with Joker than he does with other Gothamites – including the ones in costume.

As such, that one panel which, I’ve suggested, shows Batman looking into a mirror and seeing Joker – even if it’s just Bolland’s way of ensuring continuity, rather than a deliberate thematic choice – could arguably read as the most important image in the entire story.

Yet, now I’ve engaged with the alternative interpretations by Darius and Morrison – even if I don’t find them persuasive on every count – the final scene gains another layer for me, another level of possibility. I can’t help but wonder whether somehow, on some unconscious level, either Moore or Bolland must have meant that panel to be significant, and whether, all this time, I could conceivably have been reading it ‘the wrong way’.

Of course, there is no wrong way. That’s the beauty of it. What these interpretations offer us is another potentially-right way to add to the other possible ways of reading. And by identifying a potential ambiguity, even if they’re imposing it on the text in contradiction to Moore’s original intentions, people like Darius and Morrison are transforming The Killing Joke into an even better, richer and more interesting book.