INTERVIEW - Peter Watts on consciousness, first person narratives and how the future belongs to sociopaths

The single best thing about following a scene, any scene, closely is being in a position to uncover a diamond in the rough.  These days the dark arts of the PR man are so sophisticated that it is easy to forget how it feels to uncover something great simply because other people with the same tastes as you are talking about it.

Peter Watts' Blindsight is a novel that brought this simple truth home to me.  Launched at the end of 2006 by Tor with little fanfare or publicity, Blindsight immediately started generating a buzz from SF authors and critics.  When 2006 ended, the novel unsurprisingly found its way onto a number of "Best of the Year" lists.  Being a Johnny-come-lately sort, I decided to invest in the book and it prompted me to write one of the most positive reviews that I've ever produced.

Wanting to learn more about the man as well as the ideas behind the book, I tracked down Peter Watts and asked him a few questions about consciousness, first person narratives and how the future belongs to psychopaths.

Jonathan McCalmont : Peter, when you first started work on Blindsight, which came first, the Vampires and the Aliens that embodied your ideas about consciousness or the ideas themselves?

Peter Watts : They arrived in parallel, not in series.  The vampire wank dates from 2001 at least, and while the nature-of-consciousness riff started bugging me back in the early nineties, it didn't have any kind of plot to hang itself off of until the turn of the century.  I'd *always* wanted to do unconventional aliens-- who doesn't?-- but my biological training made that tough.  Every time I came up with something truly cruel and unusual, I'd barely have it written down before my postgraduate daemon chipped in with some biochemical or evolutionary reason why it could never work.  Expertise is a straitjacket, you know.  Hamstrings the imagination.

Anyway, I had three separate compartments, each bubbling along independent of the others.  And when I finally decided to sit down and try writing Blindsight, the walls between those compartments dissolved spontaneously and everything kind of mixed together.  That was good insofar as it allowed me to write a book jam-packed with cool stuff.  It was bad insofar as it only resulted in *one* book, whereas if I'd had the economic sense to keep those compartments separate I could've wrung three novels from the same amount of inspiration.

JM : I couldn't help but feel a marine biologist's disapproval behind Theseus' experimentation on the Scramblers.  Would you say that, if true, your views on intelligence and consciousness have moral repercussions on how we treat non-conscious but intelligent animals?

PW : It's a good question, but the way you asked it is a bit problematic.  The role of sf is to ask "what if", not to say "this is"; and while I did write Blindsight with a specific punchline, I really, really hope it's wrong (although I've yet to see a compelling argument why it is).  The book's views are not necessarily my views.  (They might be my *fears*, however.)

That aside, the easy answer is to say, Yes:  if an organism is nonconscious, then regardless of intelligence it cannot "suffer" by definition, so morality doesn't apply:  you might as well do anything you want to it.   But the underlying premise in your question is that there *is* some kind of objective morality, to be applied or not depending on the circumstance.

I can't argue for that on purely rational grounds.  "Good" and "evil" seem to serve primarily as tags to connote what a society approves or disapproves of at any time, irrelevant in any fitness sense except as a means of social control.   Those hardwired moral imperatives that *do* seem to transcend culture -- mother love, xenophobia, revenge/justice--  are, as far as I can tell, just tarted-up manifestations of kin-selection or reciprocal (i.e., faux) altruism.  True altruists tend to lose out over time, since by definition they spend their own energy furthering someone else's agenda;  the saint on the Titanic who gives away his life jacket will leave behind fewer genes than the selfish prick who keeps his.  (By the same token, it's very much in one's interest to promote altruistic behaviour in *others*-- which would explain the persistence of western religious institutions, among other things.)  We should not be surprised that sociopaths routinely rise to the top of law, medicine, and politics.  Corporations themselves, as legal entities, meet all the criteria for clinical sociopathy.  The success of nations has far more to do with the size of their arsenals than with any yardstick of "moral" behaviour-- in fact, "morality" is routinely redefined to include whatever the dominant players decide to do.   I can't even argue for basic human rights from first principles.  There's no morality in nature; there's what works, and what doesn't.  So if torture, for example, achieves your goals, and you can get away with it-- hey, why not?

And yet you were right about the subtext you detected.  For all my Nietzschean posturing, I am endlessly outraged by the abuses we visit upon the world and each other.  My blood boils in reaction to injustices both global and petty, and most people would probably agree this makes me a "good" person-- or at least an "ethical" one.  Yet it hasn't really proven that beneficial to me; in fact, it's proven downright detrimental.  I tend to not back down, which does not make me popular with cops, border guards, or publishers.  I came within a hairsbreadth of blowing off the contract for this very book,  in protest over a relatively minor injustice inflicted on someone *else*.  Some folks quite legitimately regard me as downright self-destructive, because I keep doing things that are not in my own best interests.  Why do I do these things?  Because my sense of fairness demands it.  Because it seems the right thing to do.

But I'm an idiot.  Life isn't remotely fair, and behaving as though it is makes about as much sense as denying gravity.  Altruism does not promote fitness.  The future belongs to the sociopaths.  A biologist of all people should know this.

Finally, I'd like to clarify my own take on "non-conscious but intelligent animals".  In the context of the book, this is reasonably clear-cut-- not so in real life, where thanks to Descartes and his ilk we seem to be in a state of denial about "animal consciousness".   I suspect that we have enormously underestimated the cognitive complexity of other creatures.  I don't think anyone who's shared living space with a cat would deny there's *some* kind of pernicious, forward-thinking agenda in those furry little brains.  And yes, there's a certain level of anthropomorphism in that observation, but I'd argue that anthropomorphism is the only justifiable approach to take in such cases. The neural architecture most fundamentally associated with the conscious state exists in everything from boney fish on up (invertebrates and elasmobranchs, not so much).  Much of our emotional response circuitry is pretty consistent across a wide range of mammalian species, and there is a wealth of behavioral data on everything from deception to bereavement in other species which, if observed in humans, would unquestioningly be attributed to conscious motives.  If we were talking about anything other than consciousness, scientists would unhesitatingly generalise-- parsimony suggests that similar systems displaying similar emergent behaviors probably share similar subjective states, even though we have no first-hand means of confirming that.   The only alternative is to assume that there are special rules that apply only to human beings and to nothing else on the planet, and that way lies religious retardation.

Why, then, do so many of us-- and not just the usual biowhores who front for zoos and oceanariums-- deride anthropomorphism as somehow "unscientific"?  Sure, other animals have less intricate brains than we do; but a lot of computer programs that run on a modern-day Itanium chip can also run on an old 386.

Here endeth the whine.

JM : That touches on one of the more interesting aspects of the approach that you took with Blindsight.  Greg Egan's short story "Mister Volition" examined the reductive theories of consciousness coming out of people like Dennett and Minsky but Egan embodied the theory in a human, because the theories are about us... not aliens.

However, in Blindsight you embodied the theories in aliens not humans and then set out the extent to which sentient creatures are qualitatively different to non-sentient creatures.  Is there not a tension there?  is it not strange that you would argue for a theory of mind and then stress the extent to which we are different to the creatures that perfectly fit said theory?

PW : When writing my tales, I like to be able to see a path from here to there.   I won't write a utopian novel set in 2050 because I can't see a plausible mechanism by which the global inertia of our current trajectory could be so completely reversed in so short a time.  Likewise, we already are conscious.  (I am, at least.)  That's my starting point.  I'm not going to write a novel based on the premise that we are all zombies, because I know subjectively I'm not one.

I might, however, be a parasite.  And something *else* might be nonconscious.  These are starting points from which I can proceed.

There *has* to be a tension here.  Even if I did throw away my here-to-there dictum, there would be little point in a tale that contained only one model. It's not a story without elements in conflict, and it's not a thought experiment without alternative hypotheses.   How can one possibly explore the ramifications of any idea without a contrasting background to weigh it against?  Or am I missing the point of your question?

JM : No... I think you nailed it. You've mentioned that you had to cut a few thousand words from Blindsight's final draft.  Can you tell us about some of the things (if there indeed are any) that you wish you could have kept in?

PW : You lost some character development-- both through winnowing surviving sections and entirely nuking others.  You lost some subplot that fed into The Gang's ongoing infiltration by a new persona: as it is there are only a couple of offhand references to set up that climactic reveal.  You lost a fair bit of back and forth between the crew as they tried to figure out what was going on-- my notes during those cuts read "terse to the point of confusion".  Which adds verisimilitude, I guess, since these are cutting-edge savantes whose speech would frequently be incomprehensible to the likes of us, but I certainly feel the pain of those who felt lost in places.  It wasn't supposed to be like that.

You also lost about thirty percent of the endnotes, but I've restored those in the pdf, as well as posting the original version on my website.

JM : It's interesting that you say that the confusion of the climax was not intentional because I thought it fit perfectly as there's no way Siri would be able to calmly analyse what happens at the climax.

PW : We're both right.  I wasn't talking about the inherent confusion of the climax so much as what I see as (in hindsight) insufficient set-up for the Panicky Persona that takes over the Gang during the climax.  It's not that that element was unclear;  Bates spelled it out explicitly in the moment, and Siri reminisces at greater length about it during the book's denoument.  But I wonder if the groundwork for that element was sufficient in the final draft.  I've always been impatient with authors who work their plots into a corner and then pull some arbitrary new element out of Batman's utility belt to save their asses (Tolkein did that a lot).  The first draft of Blindsight contained more set-up; little bits where Susan would start sleepwalking and there was some question of which persona was in control, or other hints that Rorschach was almost seducing her in some way.  You can see the barest shadow of those elements in the final draft:  an offhand reference to elevated oxytocin levels in the gang's bloodstream, and Sarasti's insistence that most of the Gang be sedated during the later forays into enemy territory.  But if you don't know that oxytocin is a biochemical correlate of "trust"-- which is to say, if you're part of 99.9% of the population-- then those references would mean nothing to you.

Now, as to the overall clarity of the climax-- yes, I was certainly trying to convey a sense of chaos and confusion, but I like to think that all the essential explanatory elements were there too.  Siri is the worst narrator in the world, but just because *he* may not be piecing it together, that doesn't mean that his reportage doesn't contain enough info for the *reader* to figure things out.  Siri often conveys information he doesn't understand himself; in fact, for most of the book that's pretty much all he does.  *He* was supposed to be confused.  *You* weren't.

Still, I really like your interpretation of it: deliberately-induced reader confusion as thematic plot element.  It makes the reader an active participant in the novel, it almost turns Blindsight into an interactive experience...

Yes.  Yes, on second thought, confusing the reader is *exactly* what I was going for.  Let's go with that.

JM : You say that if we were really to hear the characters in Blindsight talking, chances are we wouldn't understand them because they wouldn't limit themselves to the conceptual framework of the English language. However, the book is in English plaintext, presumably because Siri Keeton is "translating" for us.  This opens up all kinds of interesting philosophical ideas about the role of the first person narrator in fiction.  You seem to address many of these ideas explicitly in novel, in fact, the beating Siri gets at the end of the book is all about jolting him out of being the kind of narrator he is. Did you set out to intentionally make a point about first person narratives?

Hmmm.

Among other things, I set out to make a point about point-of-view.  In a nutshell, that's what sentience *is*:  point-of-view.  And I had to tell it through the prism of a narrator who didn't actually have one of his own (or at least, didn't believe he had one-- or at least, acted the way a conscious entity would if it didn't believe it had one. You never really find out.).  So right off the bat I'm dealing with the ultimate Unreliable Narrator; then I compound that by having him narrate the activities of people so far ahead of the curve that they'd be speaking gibberish even from the perspective of a narrator who *could* be trusted.  It's the classic insurmountable obstacle to writing post-singularity fiction:  the singularity is the light trip at the end of 2001.  What lies on the other side is, by definition, utterly incomprehensible to anyone on this side.  A gerbil cannot understand the workings of an iMac no matter *what* words you use.

Now, the crew of the Theseus is by no means post-human-- there would be no way to tell the story if they were-- but they're close enough to the event horizon for tidal forces to start pulling them into unnatural shapes.  And they've been assigned to study something even *more* distorted.  How the smoking christ am I supposed to convey that in a way us gerbils can get our heads around?

Subtitles.  Conventionalisations.  Rotation of complexity from eight dimensions down to four.

So to answer your question:  I didn't set out *intentionally* to make a point about first-person narratives,  I set out *inevitably* to do that.  There was no way to write Blindsight *without* doing that.  Whether I realised that off the top, or was simply dragged kicking and screaming to that realisation while writing the novel, I
really don't remember.

JM : One of the ideas that most puzzled me in the book was the idea that to a non-conscious intelligence our communication would be akin to some kind of viral spam.  I'm not sure I followed your train of thought there... surely all discourse, no matter how phatic or empty-headed contains some information.  Would the Scramblers not just see us as low signal to ratio types?

PW : Remember, we're not talking "low signal"-- nonsentient beings would interpret many of our transmissions as "zero signal".  Worse, in energetic terms, we're not just talking zero gain, we're talking net loss.  From the scramblers point of view (or whatever a nonsentient being has in place of a point of view), we had tricked them into eating a diet whose energetic value was less than the cost of digestion:  we were, quite literally, weakening them.

Now, you're right:  this isn't *much* of a threat, since all you have to do to avoid that cost is ignore the signal (which is why they stopped talking to us relatively early on).  But the presence of one (perceived) hostile act implies others, and future attacks might not be quite so easily thwarted.  So the scramblers were treating Human signals not so much as dangerous in themselves, but as evidence that more serious threats were in the offing.  Under those circumstances, preemption is a perfectly rational option.

This is a good case-in-point regarding your previous question, by the way.  I think the longer draft of Blindsight had greater clarity on this issue.  I could have easily made the point at the shorter length-- I just did that now.  But when you're flensing, it's only the words on the *screen* that disappear; the ideas in your head remain intact.  I think what happened in this case is that I cut for length, read the shortened scene, and saw my ideas reflected clearly in the words that remained even though they no longer existed there;   I was just seeing the residual verbiage sitting in my frontal lobes.

I really should've run this thing through some beta testers.

JM : With all the positive reviews and coverage it appears as though you've been taken to the collective bosom of the online SF community.  Did this surprise you and do you think that it has helped you?

It amazed me.  It still does.  Blindsight was a difficult project from start to finish-- I fired my agent on the first lap, my partner of eight years fired me (with good reason) on the last, and when all was said and done I expected the book to sink without a trace.  I can't tell you how much it means to see it treated so kindly by so many.  I'm still not quite sure how it happened.

As to whether I think it has helped me professionally, though, you're asking me to count prenatal chickens.  This "online bosom" of which you speak is soft and warm and wondrous, and it's lactating like Niagara Falls, but the milk itself may not have a very high fat content-- by which I mean it seems to emanate more from fans, bloggers, and fellow writers than from the established priesthood.  I don't think I'm anywhere on SFFWA's radar, for example.  I'm not aware of any real promotion from Tor.  Locus did put Blindsight on it's recommended reading list, but hasn't actually reviewed the book.  And while a number of people have described Blindsight as a Hugo contender,  I've yet to receive a single piece of fan mail from Japan (which is where this year's winner will be decided).  You can't vote for a book you've never heard of, and it was tough enough getting Blindsight out even to readers in North America.  I doubt very much that the Japanese sf community even knows I exist.

But you know something?  That's okay.  Given my low expectations out of the starting gate, the mere fact that Blindsight hasn't tanked qualifies as a major victory in my eyes.

JM : In another interview I read, you mentioned that you felt a bit burned out and pissed off with the life of an author.  Have your feelings on this changed and are you planning to write another novel?

There's no way to tell.  The ideas haven't stopped coming -- I've got hooks for another three or four books sitting in my frontal lobes, although I haven't developed any of them.  --  and the feedback I've received from the boots on the ground has been enormously heartening.  But those were never the issue.  The issue was the industry end of things, and I've had no contact with the industry since I made those remarks.  There's no new data; hence, there's no basis for reevaluation.

JM : Thanks a lot for your time Peter.  Blindsight is available for free download HERE but I would warmly recommend buying it anyway.  If only to prompt Peter into a reevaluation.

Comments

[A.R.Yngve]

About consciousness in animals... there is evidence in documentary footage, to suggest that other mammals (lions, hippos) have shown compassion and care for animals of other species, even when there is no apparent "selfish gain" for them.

Maybe consciousness in animals (I grok the metaphor that one piece of software can run on different hardware) is much more common than we thought?

In one of Roald Dahl's shorts, a scientist discovers that ALL life is conscious and feels pain; how would we be able to live with that??
:-S

As for the possibility of non-conscious sociopaths: take Saddam Hussein. Did he show signs of consciousness? (Seriously: Did he?) We know that he had terrible taste in art; does this indicate he was a non-conscious intelligent human...?
:-S

[Jonathan McCalmont]

Well compassion and care for others can be explained away in purely game theoretical terms (it's rational for me to trust you if I have no reason not to trust you as cooperation brings about great gains than competition), so I don't think you need to argue that animals are conscious.

I think the idea behind Blindsight's talk of psychopaths is that non-conscious animals see the world PURELY in game theoretical terms. So they'll be nice if it's in their interest but will not hesitate to beat you with a pool queue until you get detached retinas if it's in their interest.

Also I think the idea is that non-conscious entities behave as sociopaths, not that all sociopaths are non-conscious. Which would be an incredibly bold statement :-)

[A.R.Yngve]

Here's an example of the kind of "animal altruism" story I mentioned:
http://hosted.ap.org/dynamic/stories/P/PUPPY_LOVE?SITE=AP&SECTION=HOME&TEMPLATE=DEFAULT&CTIME=2007-02-14-16-49-39