Gleick: The Information

I just finished The Information: A History, a Theory, a Flood (2011), by historian author James Gleick. This past summer I read his book, Time Travel (2016), which was about time travel in fiction and in our hearts. [see Passing Time (My bad; it should have been titled Gleick: Time Travel, but I can never resist a pun.)]

If you read my post about the time travel book, you know I didn’t care for it, although I place the blame on my expectations, not the book. I do find Gleick, as I said then, “ambling, rambling, and meandering,” but I’m sure many greatly enjoy his excursions. I ended that review mentioning I’d like to read another book of his (a trend takes two data points).

The Information is that book, and I did like it more than Time Travel.

Firstly, forewarned is forearmed, so I was prepared for Gleick’s tendency to meander, and, dear Lord, does he. The book is ostensibly about information and society, with an emphasis on information theory, but it ambles into ancient linguistics, behavioral psychology, biology, and much more.

It’s fairly exhaustive (and exhausting) in its rambling. The library ebook version I read has 900 pages. A complaint I had about Time Travel was, despite his (in my opinion) overly wide net, a number of what I saw as notable things slipped through. His net is equally wide here, but I did not have the same sense of important fish slipping through.

For all of its length, I was — generally — far more engaged than with Time Travel. That said, there were a few sections I skimmed, even page-flipped, through. It might say something that I found the long Epilogue chapter among those. It seemed to me the weakest in the book (Gleick gets downright baroque), but your mileage may vary.

Secondly, I wonder if the material made a difference. Time Travel, in a real analysis, is absurd — it breaks causality. It’s a fictional topic and a fantastic one, at that.

But no one can deny we now live in an Information Age. Part of Gleick’s thesis is that we always have. Information has always been fundamental — just consider DNA. Nothing could be less fictional than information. Further, the history of our understanding of information is night-and-day less fanciful than the history of our understanding of time travel.

So maybe this is simply a better, more grounded, topic. I did feel he was a bit out of his depth in the Time Travel book. Here not so much.


Thirdly, speaking of content, this book was close to being a pop science book which I sometimes found disconcerting. Gleick would start to get into a technical aspect of a topic and then veer off before getting too deep. I had to keep reminding myself it wasn’t a science book.

Gleick certainly isn’t a scientist, although he does pretty well. It drove me a little crazy his repeated use of uppercase Π (Pi) rather than the correct lowercase π (pi), because that’s wrong referring to the famous mathematical constant. But it was the only actual error I noticed, so not bad for a historian tackling a technical topic.

Lastly, just in passing, Gleick seems coy with dates sometimes. Some of it is his meandering; a date mentioned pages ago is hard to remember once he returns to the topic. Are historians sensitive to the complaints students make about dates (they’re hard to remember and easy to look up)?

I’ve never cared much for history and have never been good with dates, but I’ve found that the more history I take in, the more I start remembering dates and sequencing events in time. For me it starts with landmark dates, for instance Einstein’s miracle year, 1905; or 1915 when he finally published his theory of General Relativity. After a while one starts to see a historical picture.

Still, it was a bit weird for me to be craving dates while reading a history book. What pod creature has taken over my mind?


Meanderings aside, the backbone of Gleick’s narrative concerns communication between humans. What we communicate, of course, is information, and the nature of this information has changed over time. (Yet, in some ways, we’ve at last returned to our ancient oral tradition.)

The first chapter, Drums That Talk, explores the talking drums of Africa. Early explorers mistook these for a kind of telegraph — that meaning was somehow encoded in the drumming. As it turns out, talking drums literally talk — they imitate, albeit crudely, human speech. Drummers are speaking to each other.

Because drum speech is crude, drum talk is extremely redundant and descriptive. One of the examples Gleick quotes:

Batoko fala fala, tokema bolo bolo, boseka woliana imaki tonkilingonda, ale nda bobila wa fole fole, asokoka l’isika koke koke.

Which he translates as:

The mats are rolled up, we feel strong, a woman came from the forest, she is in the open village, that is enough for this time.

It’s birth announcement.

To jump ahead, my typing those two quotes enacted a key information theory concept Gleick gets into later: logical depth, which is different from complexity or randomness, two other key information theory concepts.

Typing the first quote required careful attention to each letter. From my point of view, the sequence was essentially random and complex, and to me has no logical depth. I have no way to predict the next character.

But the second quote has many characters are almost redundant given the ease of predicting them. English has considerable logical depth, which allows me to predict. For instance, a three-letter word that starts with “th” — the third letter can only be “e”. That quote was easy to type.

(“If yu cn rd ths…”)


Gleick spends two chapters exploring linguistics and language. I confess skimming those chapters. At that point, I was beginning to think I wouldn’t like The Information much more than I did Time Travel.

But chapter four explores Charles Babbage, Ada Lovelace, and the rise of the machines, which I found much more engaging.

The two chapters after that get into long-distance communication, starting with the visual semaphore signaling towers of Claude Chappe in 1792. (Terry Pratchett uses the idea in his Discworld books.) Those were limited to how far one could see. Once electricity got involved, we began to develop “a nervous system for the Earth.”

With each advance, the semaphore, the telegraph, radio, the telephone, television, and finally cell phones and wireless communication, the world got smaller and faster and more complicated. Our access to information grew exponentially. It became, per Gleick’s title, a flood.

[Cue the John Naisbitt quote: “We are drowning in information, but we are starved for knowledge.”]


It’s chapter seven, Information Theory, where Gleick really digs into the main topic. This, to a great extent, is about the legacy of Claude Shannon.

Many found a key tenant hard to accept: Information is not meaning. Shannon was explicit that his work completely ignores meaning. What the bits mean isn’t relevant to their successful transmission or storage.

Information theory includes the study of entropy, randomness, and complexity, which are all related but different. As mentioned above, the idea of logical depth is also necessary when we consider potential meaning.

Another key concept is Kolmogorov complexity, which is about what is required to create a given sequence. [see Complexity and Randomness for details.]

I found nearly all of this engaging and worthwhile. (I did skim some of his digressions, though.)

The last chapters get into such topics as Wikipedia (which Gleick casts as a real-life Library of Babel — there are some parallels, but I’m not sure I entirely agree) and our passion for instant news. I found these chapters slightly less engaging, but still interesting.

One aside: Gleick mentions namespaces, which are very familiar to programmers. It’s the idea that, within some context, names must be unique. Gleick points out that (and I love this), rock bands, for instance, comprise a namespace. Boston might be the name of a city or a chowder, but in the rock music namespace, it’s a band. Other namespaces include brand names, baseball teams, and cities.


The final chapter, Epilogue, as I mentioned, didn’t do much for me. I couldn’t decide if he was making an ever thus claim or saying things have indeed become a bit of a mess. Reading the book I was struck by what felt like ever thus as written, but what I kept getting was that maybe those complaints were real and we’ve been sliding into a mess all this time.

I’ve been making that argument for over 40 years. Exhibit A, of course, is POTUS#45 and his 70+ million cultists. I’ve never been less happy about being right.

The thing about a flood is that it can’t be managed. The more information we have access to, the harder it is to find any one piece (and even harder to find accurate pieces).

I do wonder if modern life has gotten to be too much for many people to process. I wonder if our culture has, on average, outstripped our minds.


All in all, I don’t think I’m a James Gleick fan. His first books are about chaos, which I am interested in, but they’re decades out of date now (1987 and 1990).

While I do seem to be finding history a bit more interesting lately (I really enjoyed Einstein’s Dice and Schrödinger’s Cat), I just don’t care for his wide net. That’s totally on me; others may find it far more engaging; I find it unfocused.

But, still, overall I enjoyed this one and would recommend it to those interested in the topic. Some parts of it I enjoyed quite a bit.

Stay historical, my friends! Go forth and spread beauty and light.

About Wyrd Smythe

The canonical fool on the hill watching the sunset and the rotation of the planet and thinking what he imagines are large thoughts. View all posts by Wyrd Smythe

75 responses to “Gleick: The Information

  • Wyrd Smythe

    I got a kick out of this part about memes (an idea, term, and meme introduced by Richard Dawkins):

    “Truth may be a helpful quality for a meme, but it is only one among many.”

    It reminds me of the saying: “Lies go around the world while truth is still putting its boots on.” It’s a classic case of a teaspoon of sewage.

    Gleick gets into memes a bit, and that was another interesting section for me. Just think about the staying power of things like “Read my lips!” or the Mona Lisa or myriad other mental seeds planted in the collective consciousness.

  • Wyrd Smythe

    Here’s an interesting question: Is there a similarity between Gödel’s Incompleteness and Heisenberg’s Uncertainty relationship?

    It was Gregory Chaitin, a name almost as big as Shannon’s in information theory, who posed the question. It had been asked (more in passing) of Gödel before, but the great man had declined to answer it. Chaitin gnawed at it and had a chance to ask when Gödel was very old, but the latter declined the meeting on grounds of health.

    FWIW, my first response was, of course not, because Heisenberg Uncertainty is based on conjugate pairs that, in wave mechanics, are Fourier transforms of each other, so two precise values cannot simultaneously exist, whereas Gödel’s Incompleteness is based, in some sense, on the Liar Paradox.

    But on reflection, it’s possible to see the unstable flip-flopping of the Liar Paradox as a kind of conjugate pair, although it feels more metaphorical than physical to me. I can see why some might see a parallel, but so far I’m leaning towards my original answer (albeit perhaps softened a bit).

    • Wyrd Smythe

      Final answer: No, they aren’t alike.

      Gödel used Cantor’s diagonalization proof, so Incompleteness (as with Turing’s Halting proof) is really about uncountable sets. Gödel showed it’s not possible to enumerate all possible true statements, just like Turing showed it’s not possible to enumerate all possible algorithms that halt. Both proofs depend on Cantor’s proof that it’s not possible to enumerate the real numbers.

      The Liar Paradox is part of that proof, but it’s not the core. The core is uncountability.

      That has nothing to do with the reason conjugate pairs are mutually unresolvable with precision. The Uncertainty relationship is based on Fourier transforms in wave mechanics, and has nothing to do with enumeration, let alone uncountability.

      So, no, I don’t think they’re alike, but I can see why they might seem alike.

  • Wyrd Smythe

    Consider this bit pattern:


    Doesn’t seem random (although it could be). Does seem to be a simple repeating pattern. Does it continue; is it binary wallpaper? Does it have actual information content at all? Does it have meaning?

    What if I gave you this hint: 6×9? (Base 13!)

    What if I said it had Ultimate Meaning? 😉

  • SelfAwarePatterns

    This book sounds like it has a lot of stuff I’d be interested in, but that meandering and baroque thing scares me off. As I’ve gotten older, I’m become increasingly impatient with material that unnecessarily takes too long to get to the damn point.

    Interesting point about Shannon. A lot of people take his version of information as the scientific statement on it, but from what I’ve read, that’s a mistake. He definitely did groundbreaking work on the transmission of information, and that matters for a lot of things. But for a primal understanding of information, I don’t think he’s that good a source.

    I’ve always enjoyed history. I think you can learn a lot about human psychology from reading quality historical accounts. The versions we learned in school were usually drained of the drama and context. of the humanity, and so were little more than acknowledgement of events. (Less chance of offending school boards or parent groups that way.) I can understand why most people were turned off by it. School history also has a tendency to give us a false impression of the past. It gives the impression that the founding parents were demi-gods who carefully and rationally thought everything out, instead of deal-making politicians with a variety of human failings.

    White washed history makes contemporary times seem uniquely blighted. But when you know that a 19th century US senator beat a colleague with a cane, survived a subsequent censor vote but resigned anyway, and was then reelected back by his state, our current travails don’t seem that far out of band.

    • Wyrd Smythe

      I think [A] you’d find the information about information worth reading, but [B] like me, would find the book filled with what seems unnecessary noise. Or maybe not; you do like history more than I do, and it’s possible you’d enjoy most or all of the excursions. The really baroque writing was in the final chapter, with just the occasional flourish otherwise. I suppose historians see it as making the topic less dry.

      Shannon was kind of an Einstein in information science, but, yeah, totally; neither scientist was the only contributor even to their own fields. They were merely giants. 🙂

      (In fact, I’ve always been slightly askance at how entropy is used with regard to information.)

      We learn human psychology from everything we do (I got a lot of it, along with my early science, from science fiction), but flipping that around, as you say, it makes learning history a lot more interesting. I think I may have an attitude about history similar to yours about math: I find it interesting to the extent it aligns with and illuminates my interests. I do enjoy reading about the history of quantum mechanics, but wouldn’t so much, say, the British Monarchy.

      Sadly, as you also say, we usually teach history as badly as we teach math. It’s such a pity we don’t focus more on educating those who educate. We have little respect for the process or those who do it, which is a disgrace.

      We’ve certainly had political and social aberrations throughout history, but I wonder if a case can’t be made about scale. 70+ million people voted for four more years after seeing four. The level of widespread social delusion seems without parallel simply in terms of numbers. Or lack of agreement on facts isn’t something society has dealt with since Medieval times (other than in totalitarian regimes).

      [BTW: Had the classic admin UX suddenly change earlier today. Could no longer access classic lists or the classic editor. Got on chat with WP PDQ. Was told classic was going away. Complained and whined. Guy made me a short video about using the new editor, which was nice, but I already knew. Then, after the chat, refreshed my browser and the UX was back to normal. Still is. Not sure for how long, though. I’ll be depressed if they take it away.]

      • SelfAwarePatterns

        Every since I found out about information entropy, I’ve wondered what it meant. (If anything.)

        Yeah, I was thinking about the math thing when typing my last comment. I often enjoy history for its own sake, similar to the way you enjoy math. Although I’m not sure the history of the British monarchy would interest me all that much, at least not the recent history. (Since they ceased to be the real head of government.) The history of the early English monarchy is kind of interesting though.

        On current social delusions, I read an article you might find interesting, which puts forth an interesting theory of why Trump voters say the things they do. In essence, it probably comes down to tribal signaling, buying into unreasonable propositions to signal their team alliance. In that sense, the very fact that they receive our derision for it, that it’s costly, becomes an important part of the signaling. It’s like large gang tattoos.

        Yikes! The whole classic admin UI went away? I hope that doesn’t happen. I still use it heavily. There’s still stuff they don’t have in the new one yet, not to mention features they never ported over, like the comment search.

        I’m gradually getting used to the block editor. The one thing I continue to find very annoying is the stupid floating toolbar, which somehow always manages to be in my way. On the plus side, since I’ve been using it, the issue with my posts not showing up in the Reader has completely gone away.

      • Wyrd Smythe

        The Wiki disambiguation page for Entropy seems instructive to me — entropy is one of those ideas that’s worked its way into a lot of areas. Thing that strikes me about that disambiguation page, compared to most, is that all the links are to similar uses rather than diverse things that happen to have that name.

        I guess the information entropy concept has had value, but it’s always seemed weird to me to call it entropy. Obviously I’m not the only one! 🙂

        I don’t doubt there is membership signalling, that much makes perfect sense. Group members do signal membership in various ways (e.g. tee-shirts, pins, bumper stickers, etc). The content and structure of their beliefs in this case concerns me, though. As I’ve mentioned, I can wrap my head around the anger and sense of disenfranchisement that led to supporting him in 2016, but after seeing him in action for four years, being unable to see him for what he is… is disturbing.

        (My old friend could only acknowledge the word I’ve heard so many use: P45 is a “buffoon.” He’s vastly more than a mere clown — he and his minions have damaged this country horribly. The willful blindness involved is what stuns me.)

        The classic UX seems still in place today. I’m wondering if they activated the change and suddenly got a lot of feedback? If we both want, even need, the classic UX, maybe we’re not alone? I opened a chat window almost immediately, and the response was slow. Maybe there were a lot of chats going?

        I suppose I’m going to have to start using the block editor. The writing seems on the wall. (I worry about ever having to edit an old post. I did that on my programming blog and the block editor broke the post because the block editor doesn’t support the embedded source code feature.)

        (I have been asking myself lately why I even do this when my opinions obviously aren’t valued or really even heard. I have been having a hard time answering. I have, in many regards, “left my scrawl on the internet wall” which was always the main goal. Maybe it’s time to hang it up.)

      • Wyrd Smythe

        It wasn’t that the classic UX went away, but it changed in (at least) three ways: Firstly, the menubar on the left got much wider, which messed up my window size. Lists were scrunched and the edit window was narrower. Secondly, any link on the menu bar led to a new UX page. For instance, clicking Posts took you to the new listing, not the classic one. Which meant loss of access to the [classic editor] links in those classic listings. Thirdly, the stats window showed an error and “Page not found”.

        I kinda freaked out. Then got very depressed during the chat when the guy said the classic editor was discontinued.

        I was very glad to see the classic admin (and editor) come back shortly after that chat. Might otherwise have given up blogging then and there given how I was feeling.

      • SelfAwarePatterns

        20+ years ago when I in a department that was struggling financially. The executive my boss reported to kept making decisions that favored other departments and hurt us. In the midst of yet another decision that would hobble us financially, the exec told me that we needed to take the global view of what was “good for the institution”. I told him that we had heard that several times, and every time it meant we lost. When was it going to be our turn? He didn’t have an answer. I gave my notice a week or two later.

        That’s the perspective I fear many Trump voters have. The good of the country is meaning less and less to them as they don’t hear their good included. It’s a serious problem because the world is changing, and hearing they need to retool is bitter medicine. It makes them vulnerable to the sweet seeming lie that things can be put back the way they once were, a vulnerability Trump exploited, and I fear he won’t be the last.

        I’ve edited an old post or two in the block editor and it went okay. The content came up in the “classic block”. It’s not really the same experience, (they obviously don’t want to encourage use of it) but it was close enough for a quick edit. But I don’t have embedded code anywhere. And I have heard horror stories of old posts getting hopelessly mangled.

        On hanging it up, I hope you don’t. But if you’re not enjoying the conversation anymore, I can understand it. I gave up online discussions many years ago and came back after a long break. I had different expectations when I returned. I came back to see what arguments others could make against my reasoning, and understood that persuasion, to whatever extent it might happen, is almost always a long term process. All we can is plant the seeds. It’s completely up to the other person whether they take root.

        I don’t like the sound of those UI changes. It sounds like someone accidentally released something still in development where the plan is to redirect everything to the new UI. Ugh. I hope the feedback they’re getting convinces them it’s a bad move. We’ll see. I might have to take a fresh look at other blogging platforms. Although the last time I checked, the others all had bigger issues.

      • SelfAwarePatterns

        Apparently WordPress is changing things. This is the second Business Plan blogger I’ve seen talk about changes to their admin functionality.

      • Wyrd Smythe

        All of what you said (and more) easily explains 2016. It’s 2020 that blows my mind. It’s a testament to how far down the rabbit hole we’ve gone.

        In various ways I’ve been online since the 1980s, and it’s not a case of then being so different as much as that scope thing I mentioned. And lately I wonder if the general polarization of society has infected online discussions. It used to be possible to have a discussion; now it always seems a polarized debate. (I no longer hear the idea of, “Yeah, I see what you mean, but I don’t see it that way, because X, Y, Z.”)

        For me, at least, it’s never been a matter of persuading (although that’s certainly nice), but of hearing and being heard. I can live with disagreement, but not being heard makes the exercise pointless. It stops being about reasoned arguments. That long (so pointless) “debate” with Wysong about the BU on your post about it is an excellent illustration. (I’m still not convinced I wasn’t being trolled; the behavior is almost classic.)

        As for the WP changes, I don’t see much choice. I agree WP seems the least worst platform (although I’ve never made a study of it). I dunno; I’ll have to see what happens. I was sorta hoping to go at least ten years since that’s getting close (next July 4). I’ve quit before but I do like having an outlet for my expressive and creative urges. Maybe what I need is to change things up; find a new mode.

      • SelfAwarePatterns

        I’ve actually always found online discussions polarized, although the dimension of the polarization varies by topic. In the old days, it was between Apple, Atari, Commodore, etc. Later between Windows, Mac, and Linux.

        When I came back, I was in discussions between believers and atheists. I was shocked to discover how polarized agnostics vs atheists were. And as I gravitated toward consciousness discussions, the clashes between idealism, panpsychism, dualism, and materialism seemed endless. And of course, anything political instantly snaps everyone into left vs right.

        I know what you mean about being heard. I do hear you, although how much depends on the topic, more on scientific or science fiction, less on math (which I tend to skim), and not much on baseball (not my thing).

        I wondered multiple times in that block universe thread whether I should just shut the conversation down. But I really hate intervening between adults as long as they’re still willing to talk with each other. I think I’ve mentioned before that one of the things I try to pay attention to in a conversation I’m in is whether it’s still productive. If the same points keep getting made over and over, I try to just move on.

        On finding a new mode, have you considered Twitter? You do a lot of short form commentary in your threads. You might find a Twitter thread a natural outlet. (Warning: you will definitely get trolled there, although the block feature works pretty well.)

        On a different note, Eric Schwitzgebel did a post a while back which has had me thinking about the way I approach things. Maybe it would give you ideas.

      • Wyrd Smythe

        People certainly have always had their allegiances. My sense is that we’ve gotten more violent about disagreement and less willing to admit any viability to opposing positions. (An awful lot of public people get death threats routinely.) Per the post you linked to, everyone is being very closed. I think people used to be more open. (Admittedly, I hung with a pretty technical and liberal crowd. High-tech hippies, basically. Pretty open in their philosophies. 😀 )

        Baseball and math posts I wouldn’t expect everyone to listen to. It’s when I’m in a discussion with someone and I realize they don’t seem to be hearing me. In some cases it feels I’m not being understood but the other person isn’t willing to stop me and ask WTF I’m talking about. In other cases, it feels more like being ignored for whatever reason. Either way, it’s hard to move anything forward when there’s no traction.

        A tactic I use is reflecting back to show understanding. I look for it, too, that echo that indicates I’ve been heard and understood. (As I’ve mentioned, it’s fine if the next words are, “I don’t agree, because…” At least I know I was understood.)

        I’ve managed to avoid Twitter so far (and deleted my Facebook page many years ago), although I have considered it. I’ve also thought of just doing some really short “Tweet-like” posts. I would really like to (one way or another) get rid of my piles of notes and just blog more in the moment. Those posts are always easier to write, anyway. Some of the others take a lot of research, thought, and diagrams (and I’m getting so tired of making diagrams).

        FWIW, a guiding principle of mine regarding being open-minded is that an open mind is a good thing to maintain, but not so open your brain falls out. I’ve never believed it’s the case that “anything is possible.” 🙂

      • Wyrd Smythe

        As an aside about the socio-political scene and being open-minded, a favorite book of mind is Idiot America: How Stupidity Became a Virtue in the Land of the Free (2009), by Charles Pierce. A key thesis of which is that we used to be a lot better at sifting the wheat from the chaff when it came to craziness.

        Americans have always been a little, um, different, but, Pierce argues, we used to be better at shining on the 99% crazy and finding the 1% gold nugget off-kilter minds sometimes turn up. Those are those nuggets that change things. (It’s part of why I don’t necessarily give the provenance of an idea the weight many do. The best minds can be crazy, and the crazy minds sometimes see a truth others missed.)

        (That said, I think a strong consensus model of reality that demands extraordinary proof for extraordinary claims is a good thing.)

      • SelfAwarePatterns

        Sounds like you hung with an open minded bunch. Where I live, I’m more used to fairly closed minded attitudes. Most of my friends, growing up, followed the more standard human condition and were suspicious of new ideas. My dad was an exception, at least in my younger years. So I guess I’m just more used to people being dug in on their positions.

        In conversations, often people are imprecise in what they say, so I often respond with something like: if you mean X, then here’s my response, but if you mean Y, then my response is this. It’s not unusual for people to reply that none of those are what they meant and then try to explain again. The problem is people often seem to settle on stock phrases they think very clearly communicate a point, and keep repeating it when questioned. (I’m guilty of it myself, although I try to recognize when I’m doing it.)

        The broader issue is that communicating clearly on many philosophical and scientific topics is a challenge. Often the available language doesn’t do a good job of conveying a concept. And people often come to concepts with different conceptions that get in the way or receiving others.

        Twitter is definitely not for everyone. In truth, I mostly just share stuff on it and lurk. The conversations can be interesting, but they’re ephemeral (particularly with the new “fleet” feature they recently introduced) which I’m not wild about. I still find blogging more congenial.

        There’s a lot to be said for short posts. People have a higher tendency to read them all the way through. And some of my highest engagement posts have been short almost throwaways. I say that as someone who often finds it painful to keep thing short.

        On being open minded, one trick I try to use is never to allow myself to reject something simply because it seems ridiculous. I need to have logical reasons, or even better, contradicting evidence, to reject it.

        Of course, it’s impossible to do that for every proposition that comes up. Unfortunately, figuring out which propositions to sink time into usually involves considering its source, or how much support its garnering among relevant experts. So, often my logical reason is that scientists don’t seem to find it plausible. I realize that’s a very imperfect heuristic. Sometimes wisdom does come out of the mouth of babes. The problem is the vast majority of what comes out of the mouth of babes is babble.

      • Wyrd Smythe

        Yeah, born in New York City, almost 20 years in Los Angeles, and even here in Minnesota, the Twin Cities are a hot-bed of progressive liberalism (to the dismay of most of the rest of the state). 😉

        I quite agree about people being imprecise (not just in language but in thinking), and that dual response is a good tactic. So is questioning them, although, as you mention, sometimes it just leads to restating. (Almost like the gag about Americans speaking more loudly to foreigners who don’t speak English, as if that somehow works better.)

        As you say, it’s a challenge, and I suspect that may be part of what leads to my long comments (and posts). Communicating a complex topic involves a lot of nuance. A deep understanding of something always seems to involve a lot of details. That canonical answer to so many technical questions, “It depends on what you mean…”

        Which is why short posts tend to be few and far between for me. Stuff I’m into usually gets complicated. (Geeze, even baseball turned out to be hugely complicated and nuanced; arguably the most so of any sport. Figures I’d find it fascinating.)

        The thing about an idea seeming ridiculous is that my ridiculous is a logical judgement. (At least when we’re talking about science and philosophy.) The reason an idea would seem ridiculous to me is that it offends my sense of logic and physical reality.

        There is, I think, an interesting bias in our culture that our judgements come (solely) from our emotions and (presumed irrational) beliefs. There is much truth to that, but the goal of the intentional mind is that judgements come from intellect, logic, the dialectic; whereas beliefs come from judgements. We’re only human, of course, but that’s the goal of an intellectual. (I’ve long loved the Albert Camus quote, “An intellectual is someone whose mind watches itself.” Indeed.)

        (In fact, as I’ve ranted about often, our culture has gotten itself into a mess by giving so much primacy to emotions and beliefs over intellect and logic. The dialectic in our culture doesn’t really exist anymore.)

        The thing about plausible is that reasonable intelligent people can see it differently, even among the learned. (MWI and computationalism are good examples — even expert opinions vary about what is plausible.) I’m finding I’m less and less interested in what might be in favor of what is right now. Improving my math and physics is challenging but rewarding. I’m actually starting to understand the Schrödinger equation and QM. After the last four insane years of uncertainty, it’s nice dealing with something concrete (albeit having Uncertainty all its own 🙂 ).

      • SelfAwarePatterns

        On ridiculous ideas, I get what you’re saying. But it’s very easy for me to just assure myself I’ve arrived at that conclusion logically. By forcing myself to justify it with the logical details, in essence to prove it, I make it less likely I’m fooling myself (or anyone else).

        I’ve also noticed that doing that often lowers the temperature of the discussion. I’m not calling someone’s idea silly (or by implication, them). I’m laying out obstacles to it, which gives them a chance to address them. (Not that I’m a saint of consistency on this.)

        On judgments coming from emotion or intellect, I think it’s extraordinarily difficult to be sure we’re really reaching it through intellect. It’s often very easy to see when others are actually engaging in motivated reasoning, but very difficult to detect it in ourselves. History shows even the most brilliant minds can fall into that trap.

        The only way I’ve found to test it is to put ideas out there and see what arguments people can make against them. We all have blind spots, but hopefully others have different ones than I do. It’s far from perfect, but better than nothing.

        Good to hear you’re making progress on Schrödinger. That’d be a post I’d be interested in!

      • Wyrd Smythe

        We’re talking about the same thing here! The goal of self-awareness is closing that gap you mentioned between seeing others and seeing oneself. It’s a matter of turning that same analytical spotlight on oneself.

        It does require learning to be rather harsh and unforgiving with oneself. (Like editing means learning to delete one’s precious words!) I have a number of distinct voices in my head. They’re all me, of course; I’m not fragmented or confused about my identity. But in a sense, I carry my worst critics in my head — they ask much harder questions than people on the internet ever do, because they know fully the weak parts of my thinking. 🙂

        It’s an intentional approach. One creates, maintains, and to listens to, those self-agents. It’s learning to give oneself the same critical analysis one would give anyone else. In that context, I think it is possible to separate emotional views from logical ones. Not always, and anyone can fall prey to mental errors, but an intentional mental discipline mitigates it a lot. It’s what that Camus quote is talking about — the mind watching itself.

        That said, definitely other eyes see things one misses! Living in a vacuum (or a bubble) is a good way to diverge into mental error. (I very much like the metaphor of the Johari Window.)

        Getting into the Schrödinger equation has raised a new question I need to ask a trained physicist. The SE describes particles, or ensembles of particles, but does not describe particle creation or annihilation (quantum field theory does). That seems to raise an issue with regard to a wave-function describing anything but the simplest quantum system. So now I’m even more puzzled about how to use the SE to describe any of the experiments (two-slit, beam-splitter, spin measurements, etc) because, as far as I can see, any measurement involves particle annihilation. (Even reflection by a half-silver mirror, I believe, involves an electron absorbing a photon and then emitting a new one.)

        (I’m starting to look into spontaneous collapse theories, because I’m convinced a huge key to the puzzle is figuring out what divides the quantum realm from the classical one.)

      • SelfAwarePatterns

        On asking a physicist, one option is the physicist at Ask a Mathematician / Ask a Physicist. He just recently did a post answering a couple of questions about the MWI, including the energy question that’s been bugging me. (Never occurred to me to me to submit the question to him.) Turns out energy dilution is not an issue. Or at least, he notes, the math for it works. He finishes with a plug for RQM (his favorite interpretation).

        The nice thing about spontaneous collapse theories is they’re supposed to make testable predictions. Although from what I understand, the propensity to collapse remains something of a free parameter that can keep being extended as larger and larger systems are held in superposition. But my biggest issue with collapse theories is the collapse itself.

      • Wyrd Smythe

        I read the post, but I’m going to have to read it again. I’m dubious about the account of the two-slit experiment, and I don’t see where he gives a good answer to the energy question. At one point he writes, “Evidently energy and existence in general is the same way: perpetually normalized.” This apparently proceeds from the bit about odds, which I don’t see how it applies. Also, “evidently”? And I’m not sure what to make of the idea of existence or energy being normalized.

        The thing about the two-slit experiment, which we discussed a while ago and seemed to agree about, is that it’s not clear multiple worlds are involved in the photon’s flight and interference. It could be the case multiple worlds are involved where the photon lands, since that does involve multiple outcomes of a quantum event.

        But I don’t think the superposition of the photon going through both slits involves multiple worlds. I think it’s a superposition in this (one) world. So his explanation involving the two-slit experiment has me wondering.

        And I’m not sure what you’re seeing as a solution to the energy question. (I noticed that his next line after the one I quoted above is, “If not then you should be able to tell the difference pretty quick.” To which I thought, “Yeah, exactly, and we don’t, so doesn’t that falsify the MWI?” 🙂 )

        I’ll have to read it again. Maybe I can make more sense of it.

        Yes, I like the notion spontaneous collapse theories can be tested. I was reading a paper recently about what seems like a definitive test (or so the authors claim) using equipment foreseeable in the next few years.

        They are characterized by two parameters, λ and r_C, the collapse rate and the correlation radius. The basic idea is introducing stochastic collapse due to noise.

        I don’t quite follow your last sentence. In general collapse is a huge question, but collapse theories seek to explain it mathematically, so why still a problem with collapse?

      • SelfAwarePatterns

        On the energy thing, my main takeaway is that my concerns about some quanta of energy in that context appeared to have been misguided. Energy can be diluted indefinitely. Under the MWI, all our observed energy levels relevant to each other would constantly be renormalized, which is why we don’t notice and can’t detect any difference.

        On the two list experiment, talking about when we have one vs multiples worlds is not absolute event. The closest is probably when decoherence happens, so when the quantum system is no longer isolated and information about it leaks into the environment, in other words, when it has wide ranging causal effects and the environment in turn has effects on it, in other other words, when it becomes entangled with the environment.

        Of course, “the environment” is a relative thing. To Wigner’s friend in an isolated lab, that lab where they conduct the measurement is the environment, but to Wigner sitting outside the lab, the environment is where he’s sitting. To Wigner’s friend, the collapse / world split happened when he measured his quantum system. To Wigner, it didn’t happen until he opened the lab door and asked for the results.

        On collapse, my understanding is that objective collapse theories have mathematics on when the collapse might happen. So they improve on Copenhagen’s vague language about the distinction between the classical macroscopic world and the quantum one. But I haven’t seen anything where they fundamentally get into what the collapse might be.

        To me, the collapse as a concept only make sense epistemically. In that sense, I think Bohr’s language was meant to be explicitly vague so as to make clear it was about our interaction with the quantum system, not the system itself. The collapse as a postulate to describe actual reality, I think, raises more questions than it answers. It reifies indeterminism, non-locality, and other quantum weirdness.

        (RQM keeps the collapse but promises to preserve locality. I haven’t been able to find a cogent explanation on how that is. Rovelli has a book coming out next year on RQM, which will hopefully shed some light no this.)

      • Wyrd Smythe

        I hear the assertion that ‘energy dilutes’ but I’m not hearing an explanation of what seems an extraordinary claim. It isn’t just E=mc^2; what about Planck’s constant? (Which is in Joules.) Is that really changing all the time? What about α (the fine structure constant)? What about gravity, which depends on mass/energy? Why doesn’t gravity get weaker?

        When we talked about the two-slit experiment recently, I was under the impression we agreed branching, if it occurs at all per the MWI, involves (only) where the photon landed? (The detector screen can be realized as a fine grid of separate detectors making photon detection similar to a beam-splitter experiment.) But I can’t parse your comment about two-slit experiments; can you be more concrete about what you think happens?

        The Wigner thing depends on the belief that macro systems can be in superposition, which I also see as an extraordinary claim. (I’ve mentioned the extreme fragility of quantum states.) I need convincing the cat is superposed, let alone the experimenter or Wigner. (And I need physics on how decoherence allows matter to coincide.)

        Objective collapse theories, I believe, seek to explain why as well as when, but I need to explore them more. I want to find out if they might account for the quantum/classic divide (which I think is real and why cats aren’t superposed).

        Treating collapse as epistemic makes perfect sense if the SE is likewise considered epistemic, and I’m beginning to think we have to. If it’s right that the SE doesn’t account for particle creation or annihilation, then reality can’t possibly be the SE wave-function. That’s not necessarily to discount the MWI, just that it can’t depend only on the SE — something more has to be happening.

        I haven’t looked into RQM except in passing. (I see Rovelli as something of a fantasist, which isn’t encouraging. I’ve been underwhelmed by his books.)

      • SelfAwarePatterns

        On energy dilution, a joule is a unit derived from other things. If all of them have been proportionally diluted, then I think it would change in the overall quantum universe, but not in any emergent classical one. That MWI FAQ mentions that the MWI predicts that spacetime must be quantum, and so branching. I think that would handle gravity (and related relationships) being diluted.

        On the double slit experiment and MWI branching, if I gave you that impression, I might have expressed myself poorly. Remember, prior to a point where we decide there’s been a “split”, the branches already exist, even if we just refer to them as “states”. It’s just a matter of how far they’ve spread.

        I do think decoherence is a natural event to speak of a “split”, since it’s where the interference effects become undetectable. All that’s required is information about the particle/wave spreading into the environment. That can be when it hits the screen. Or it can be at the slit if there’s a detector. But as I noted above, “the environment” is a relative thing.

        Certainly quantum states are fragile from our perspective. But the question is, what happens when they appear to change to classical ones? Nothing about the fragility prevents macroscopic superpositions, only our ability to measure or make use of their interference effects.

        With that in mind, what about a macroscopic superposition makes it a more extraordinary claim than the instantaneous annihilation of most of the wave function, across all spacetime intervals? Both seem extraordinary, but one is an extrapolation of the raw quantum formalism, while the other is a postulate.

        If objective collapse theories do get into the why (and how), I would find them more interesting. My understanding of them (admittedly from popular accounts) is that they’re simply introducing additional postulates for an ontological collapse. But if they attempt to dissect the collapse, that would get my attention.

        I could see the SE being partly epistemic. But it seems like at least some aspects of it have to be modeling something real. If not, where are the interference effects coming from? A thing having causal effects is enough for me to see it as real in some sense. From what I’ve read, objective collapse theories do see the SE as at least partly ontological; otherwise what is ontologically collapsing?

        Yeah, I have similar concerns about Rovelli. He challenges assumptions, which is good, but his alternative assumptions don’t strike me as that compelling. But there are enough physicists who find his ideas viable that I feel the need to learn more about them. I just hope his book has more meat than the stuff I read before (but not so much that it’s unreadable for a non-physicist).

      • Wyrd Smythe

        We’re going down the MWI rabbit hole again, and I’m fine with that, but I’d like to keep this as much about specifics and details as possible. This thread started, in part, over the value of the dialectic, and this is a good chance to test that. If you’re not into getting specific and detailed, let’s talk about something else.

        re joules. You’re right they’re a derived unit — expressed in newton-meters and seconds, and newtons are expressed in kilograms, so at root, we’re talking mass, time, and length. That doesn’t change the issue. Since I doubt time or length is changing due to branching, it has to be mass. (The joule is also defined as “the energy dissipated as heat when an electric current of one ampere passes through a resistance of one ohm for one second.” This implies that current or heat must also be affected.) Gravity depends on mass, so if mass is being as thinned out as the MWI suggests, wouldn’t gravity be weaker?

        re two-slit experiment: Do I understand you subscribe to the idea of a universal wave-function that contains all worlds and branching is just a case where two wave-functions that have been thus far identical (for 13.8 billion years) finally branch?

        Can you give a precise account of what happens in a two-slit experiment? Exactly why we get the results we get? I’m still not following what you’re seeing there. Where, and when, does decoherence apply? After sending the first particle, how many worlds are there? After sending many particles and the interference pattern exists, how many worlds are there? In both cases, what do those worlds see?

        re fragility of quantum states: I don’t follow; isn’t it precisely that fragility that prevents macroscopic superposition? We’ve never ever seen a macro object in superposition, so the extraordinary claim to me seems that they can be. Given that the classical world clearly isn’t quantum, what justifies the extrapolation from raw quantum formalism?

        The thing about collapse is that I’ve never understood the big deal. It’s something we just haven’t figured out yet, and until we do I can’t get too excited about how “impossible” it is. The complaint is that the SE doesn’t deal with it, but that seems to put the SE on an overly important footing. Maybe nature isn’t playing by those exact rules.

        Maybe, in being part of the raw formalism, it’s not the whole picture?

        There is another issue. The SE for a single particle requires three dimensions (it may require six, three for position, three for energy). In an ensemble, those dimensions are required for each particle. Further, a lecture I watched pointed out something interesting: the use of i in classical physics is helpful, but not fundamental. But if you look at the SE, i appears only on one side of the equation. So i is fundamental to quantum mechanics in a way it isn’t in classical physics.

        So the ontological wave-function has to be something with a vast number of dimensions and which uses complex numbers in a fundamental way. What can that be? It seems Tegmarkian or Platonic. If the WF is the ontological center of MWI, then what exactly is it?

        Finally, there is that the SE doesn’t handle creation and annihilation of particles, so can’t be a full description of reality. Something more is required.

        re objective collapse theories: I believe they do attempt to dissect collapse (assuming you mean explain what’s going on — why quantum states vanish). I read a paper recently about an possible experiment testing continuous spontaneous collapse (CSL), and it’s made me interesting in learning more. It seems to be along the lines I’ve been considering.

        I continue to believe that what we really need to crack the nut is an understanding of superposition and why it happens for tiny isolated things but not anything else. (Because superposition is very, very weird.)

      • SelfAwarePatterns

        I guess I brought up the MWI again by referencing that Ask a Physicist post. Sorry, wasn’t my intention to bring us down a new rabbit hole. I do my best to be precise, but it’s not possible to precise on everything. The key thing is whether you’re getting the points you want in a response. All I can do there is suggest you ask (precisely 🙂 ) if you didn’t, keeping in mind I’m just an amateur enthusiast, not a physicist. (Warning: I’m also slightly looped out on painkillers this morning after oral surgery yesterday, so apologies if anything seems side ways.)

        I do think gravity would be diluted, has to be a part of that renormalization. That’s what I meant when I said the MWI implies quantum and branching spacetime.

        I think under the MWI, the universal wave function is the central concept. The second part of your question, about two wave functions being identical and then diverging sounds like Deutsch’s conception. I’m not onboard with that one yet. I’m not saying I would never be, but it doesn’t click in my mind at this point.

        On the two slit experiment, the photon is emitted and begins spreading out as a wave. Much of it impacts the first screen, which means it interacts with it and becomes entangled with it, an entanglement that spreads throughout the screen and then the environment. leading to a branch of the environment’s superposition where the photon never made it past the first screen.

        But for the portions of the wave that do make it through the slits, those portions interfere with each other before impacting on the second screen. The second screen being some kind of photosensitive material, it leaves a mark, but each part of the wave leaves it’s own small mark, so that part of the wave becomes entangled with the back screen, in a repeat of the process with the first screen, except in this case, the mark has magnified the result leading to diverging outcomes in each branch of the environment’s superposition.

        The above paragraph is true unless a detector is placed at one of the slits. If so, then that portion of the wave triggers the detector, leaking information into the environment, while the rest of it doesn’t. That means information is cascading out at that point, which makes the portion of the wave that triggered the detector no long coherent with the rest, and the interference effects disappear.

        In most branches of the environment, the photon never makes it past the first screen. Without the detector, in a smaller portion of branches it goes through both, interferes with itself, before impacting a particular part of the screen (causing these branches to diverge from each other). With the detector, the divergence happens at the point of detection.

        Hopefully somewhere in all that is what you were looking for?

        On the fragility of quantum states, my point was that you’re assuming when they disappear to us, that they cease to exist. Under the MWI, we would observe exactly the same thing as under a collapse interpretation, the loss of interference effects. The fragility of the states appears interpretation neutral.

        I see the collapse as the central postulate of most interpretations of QM. And it’s the one that brings in most of the weirdness you see in popular articles on QM. So to me it is a big deal. Again, I have no issue with it as an epistemic acknowledgement of a brute empirical fact, that prior to measurement we have wave effects and afterward particle effects. Keeping it epistemic, apparent, phenomenological, leaves us open to various solutions (such as the MWI).

        But almost no one can keep it instrumental. As soon as we start talking about it as an ontological thing, the paradoxes start. It becomes a gigantic postulate, one that, to me, requires a lot of justification. Maybe there really is a collapse. But if so, postulating it should be acknowledged as admitting a gaping hole in our understanding.

        Maybe the SE isn’t the whole picture. But it seems like we could “maybe” ourselves forever. Do we have evidence for any additionally required formalism? (I’m asking, not rhetorizing.)

        I don’t grok the SE, so can’t discuss it in detail. But the “dimensions” you mention and that I commonly see discussed sound more like degrees of freedom, accounting dimensions, rather than implying that additional spacetime dimensions or something would be required.

        On something more than SE being required, here I hope you don’t take offence, but you’re just learning the SE. People who have studied it their entire career don’t seem to see the issues you do. (Although it is interesting that Deutsch sees matrices as more useful for some purposes.) So I think caution is warranted.

        Superposition of waves is a normal thing. Superposition of things like spin strike me as very strange, but spin already seems like a very strange thing anyway. The question is whether quantum superposition really is only microscopic or if it’s universal. A lot hinges on whether we ever find evidence for an actual cut off point. Copenhagen and its cousins say we will. MWI and its cousins say we never will. In case you haven’t seen it, there’s a recent Quanta article about a new theorem on this subject that’s worth checking out. (Although I think the article overhypes RQM.)

      • Wyrd Smythe

        Good morning! Hope you’re enjoying your pain killers! 😀

        As a light aside, since you mentioned referencing that Ask a Physicist post, in the comment you were replying to I had deliberately avoided mentioning the MWI because we both do seem to find it irresistible. But I did want to mention the SE and particle creation/annihilation, so I kind of unlocked the door, so to speak. (And certainly walked through it when you opened it. 🙂 ). Which is all fine, the topic fascinates me, and I’ve been actively studying it. The goal is to “keep it 100” as they say and focus on concrete specifics. We’re off to a really great start, I think!

        On a procedural note, in the old days quoting each other was standard, and I developed what I saw as a highly conversational style of quoting relevant bits and directly replying as if in a conversation. In the modern comment era I’ve gotten the impression this puts people off, but I still like the precision and brevity of quoting. I’ll try to keep it minimal; I hope it’s not too off-putting.

        * Gravity: If it’s diluted, wouldn’t we notice it getting weaker? In a vacuum, a feather and bowling ball fall at the same rate, determined by the Earth’s mass. Wouldn’t free-fall rate change (it’s different on other bodies)? Per GR, gravity and acceleration are equivalent, so if gravity is somehow changing, acceleration has to also change, but that’s tied to velocity changing.

        (FWIW, I have other energy questions. Radioactive decay half-life is tied to the energy of the nucleus. If energy was diluting, why doesn’t it affect radioactive decay? Another example is emission lines, which are also tied to energy. For that matter, the color of light is its energy. Why doesn’t color shift? The notion of diluted energy or thinned mass seems to raise a lot of questions. Joules may be derived, but mass is a base unit. It’s hard to see how it wouldn’t affect other properties where time or length are part of the equation.)

        I’m wondering if it might be better, at this point, to break this up into distinct topics and give each their own thread. You can choose to respond (or not!) as your interests incline. I’ll leave this thread for any discussion about gravity or energy with regard to the MWI. (Or any other remaining topic from upthread.)

      • SelfAwarePatterns

        Hopefully I’ll be able to wean myself off the stronger stuff today and just be on ibuprofen. Thinking should be sharper after that. (Or as least as sharp as it ever gets.)

        I’ve gradually lessened my use of quoting the other person in discussions after noticing that it tends to lower the conversational temperature. It’s too easy to slide into a rut of just sniping at each other, cutting the quote so it’s out of context, etc. That’s not to say I never quote anymore. There are still times where it’s useful. But I try not to make it my default mode. But I’m not that worked up if you or others do it.

        On noticing the changes in gravity and other things, I think that’s what the Physicist’s reference to renormalizing was about. If everything is equally diluted, then all the relations, the ratios, remain the same and there’s nothing to notice. A lot of this is similar to us not noticing that the Earth moves and spins. If everything is tangled up in it, we shouldn’t notice it.

      • Wyrd Smythe

        Yeah, I noticed the same chilling thing about quoting. The irony is it’s my attempt to make things conversational and responsive. I try always to do justice to the other person’s point. I like the clarity and brevity it provides, so it’s a pity.

        I assume you noticed; I started four new threads below, one per basic topic. Mix, match, ignore, or combine, as needed. Or start new threads; whatever.

        The thing about gravity is, why isn’t free-fall rate affected? That rate depends on the gravity, and is measured in terms of time and distance. How does normalization account for that?

        We don’t directly sense the Earth spins, but we surely notice something does. The sense disconnect caused most to mistake it, but we definitely noticed something was moving.

        That’s not the same with this thinning of energy/mass. Nothing we do can notice it, which, per the gravity case, seems like might be impossible. Many of our observations involve mass/energy combined with length or time, so it sure seems something should be noticeable.

  • Wyrd Smythe

    This thread is for the two-slit experiment discussion:

    * Deutsch’s conception: Yes, exactly what I was thinking. Your description makes me think I was right the first time — you do see the branching occurring due to particle interaction. (Which would be aligned with Everett’s original conception.)

    “On the two slit experiment,…”

    The whole paragraph, I think we’re in sync. Can I ask you to elaborate on what you mean by “becomes entangled” with the (first) screen and “spreads throughout” the screen and environment? What, exactly, entangles and spreads?

    Is it fair to state it as: The photon’s matter wave spreads out and potentially interacts with every available electron in the first screen (and any sides, floor, or ceiling). This potential interaction results in a distinct branch for every such available electron. There is a vast number of such branches where the photon was absorbed.

    (In contrast, in SWI, the above is the same except being absorbed just results in the photon ending its flight.)

    We’re in sync regarding the second screen, too. Again, there are a vast number of possible interactions, and this version of the MWI says they all result in branches. In this case, however, interference means the probability of where the photon interacts is affected such that an interference pattern emerges over time.

    Now here’s a critical point:

    “But for the portions of the wave that do make it through the slits, those portions interfere with each other before impacting on the second screen.”

    I agree. It’s the matter wave that interferes with itself, yes?

    Do you agree it is not the case that one branch has one photon and another branch has a other photon, each going through a different slit, and then those branches somehow coherently interfere? Further that the branches then come back together to account for the photon interacting with a given electron?

    Do we agree branching is the result of the photon interacting?

    One question at this point: All the branches where the photon is absorbed before passing through the slits, are they distinct? Or, since a lost photon is just lost and where it’s lost isn’t recorded, are those all branches distinct, or is there just one “lost photon” branch? Every emitted photon has that same vast number of possible interactions, so does shining a flashlight at a wall result in gazillions of essentially identical branches? Of if they”don’t matter” (for some definition) are they not distinct branches?

    [As an aside, I did some reading about how photons expose film, and one thing is that it apparently takes four photons to expose a single grain of silver nitrate. That may be why we don’t see much single-photon two-slit photography. (The ones I’ve seen used electrons.) Later I ran into photography again with objective collapse theories. Apparently, explaining the latent image — which can exist for many years — is one goal. It does seem to raise interesting questions in a MWI context. The film must be in superposition for a long time until developed?]

    I think we’re in sync regarding placing a detector, too. The wave phase information is lost to the detector and thus not available to interfere.

  • Wyrd Smythe

    This thread is for discussion of quantum states and collapse.

    When I say a quantum state is fragile I don’t mean to imply the information that state contains ceases to exist. What I mean is that interaction with the environment completely overwhelms it with random information. The original information is no more available to us than the wave pattern of a rock dropped into the surface of a windy lake surface. In principle the rock’s wave pattern is part of the total, but it’s so mixed with an overwhelming number of other waves as to be indiscernible.

    I quite agree that fragility is interpretation neutral.

    Collapse is definitely seen as a big deal by most. I’m an outlier in wondering what the big deal is, although I should be more precise: I’ve wondered what the big deal is in terms of the Schrödinger equation, because the complaint is the lack of math, and I can’t see not having figured out the math of something as that huge of a deal. (What I’m learning about the SE increases my sense of this, but the SE will be a separate thread.)

    What I do agree is very weird about “collapse” is the apparent non-locality and the apparent randomness. (I suspect folks have gotten over-focused on “collapse” whereas I think the focus should be “measurement” or “interaction.”)

    (Of the two, the randomness vexes me most. Why does a photon’s matter wave interact with the matter wave of a specific electron? Is there something underlying that determines the choice or is it truly random? Non-locality occurs in virtual particles and Bell’s experiments, so it an apparent fact of reality. It can’t violate causality and may speak to a deeper physical reality than the three-dimensional one we know. (It’s weird, but I take comfort from it not being useful.))

    I know the MWI answers randomness, but its answer to non-locality challenges me. Under Deutsch’s conception, a universal always-existed wave-function accounts for it, but the idea of different phases of reality that meet and either cohere or don’t raises a lot of “how does that work” questions.

    You’ve asked “what’s collapsing?” I think that’s the heart of it.

    For now, let’s assume the SE is an epistemic description of a matter wave. As such, the SE changing due to measurement is purely epistemic, so let’s focus on the putative matter wave — whatever that is.

    We know it acts like a wave up to the point it interacts. In some cases the matter wave appears (when the laser emits a photon) and then disappears (when the photon is absorbed). This happens in all interpretations, so the real question is: What is spreading out? What, really, is the nature of a “matter wave”? (I have ideas related to QFT.)

    Such a thing does seem quintessentially quantum. And talk about fragile, just looking at it makes it do the c-word. It’s very shy. 🙂

    I know you appeal to the MWI for answers to both randomness and non-locality. I understand the former, but you’d have to explain to me exactly how the latter works. I get that you think the superposition expands at light speed, but what is the exact mechanism in play when they meet?

    • Wyrd Smythe

      Maybe it’s worth unpacking some different notions of “collapse” — it has some different flavors.

      Collapse often refers to position localization. The matter wave, which has a definite momentum but isn’t localized “collapses” into a point interaction. This flavor implies non-locality in how the spread out matter wave suddenly vanishes from all points in space except the interaction point.

      The Schrödinger equation experiences this flavor of collapse as a non-linear discontinuity in an otherwise linear continuous function. The wave-function jumps from describing a definite momentum and indefinite position to a definite position (and thus indefinite momentum).

      Something that complicates this picture is that localization measurements can involve absorbing the observed particle. In such a case, Schrödinger equation no longer has a particle to describe. Energy and quantum information has been passed to some other particle’s wave-function.

      Another flavor involves measuring spin. This also causes a jump in the wave-function, in this case from indefinite spin to definite spin on the selected axis. After the measurement, note that spin on other axes is indefinite, but correlated to a degree that depends on the axis measured.

      Also note that we can make successive measurements at different angles. We can have an experiment that fires particles with unknown spin into a detector that determines vertical spin and pass the spin-up particles into a second detector that measures the horizontal axis. If we pass the spin-left particles into a third detector that measures vertical spin, we’ll get equal numbers of spin-up and spin-down particles.

      We can notate it as:

      |*⟩ ⇒[Y]⇒ |U⟩ ⇒[X]⇒ |L⟩ ⇒[Y]⇒ |U⟩+|D⟩

      At one point the spin was definitely |U⟩. The X-axis measurement giving us definitely |L⟩ destroys that former state. The second Y-axis measurement therefore returns undetermined results.

      So spin flavors of collapse are more subtle than position localization flavors.

  • Wyrd Smythe

    This thread is for the Schrödinger equation …

    “Maybe the SE isn’t the whole picture. But it seems like we could “maybe” ourselves forever. Do we have evidence for any additionally required formalism? (I’m asking, not rhetorizing.)”

    You don’t sound convinced! 🙂 The SE describes existing particles and, as far as I know, does not describe particle creation or annihilation.

    More than one lecture I’ve watched refers to it as the F=ma of quantum mechanics. In the same way that F=ma describes the motion of classical particles — allowing us to predict where they’ll go or where they came from — the SE describes the motion of quantum particles. It likewise allows evolution forwards and backwards, but rather than providing specific answers, it provides probabilities. That’s the quantum difference — probabilities, not certainties (because waves).

    So it makes sense it isn’t a complete description of reality any more than F=ma is. There’s more to the picture on both the classical and quantum levels. In particular, it’s quantum field theory that describes how particles are created, exist, and are annihilated.

    (BTW, this is why I’ve been saying for a while now that MWI should decouple itself from the SE and focus more on the physicality of what might be going on.)

    Speaking of physicality, yes to degrees of freedom (each particle has its own set). Note that, for each particle the degrees are physical 3D space, X, Y, Z. There is no implication the particles are doing anything extra-dimensional.

    But the wave-function describing them all is comprised of all those dimensions. Ultimately, it’s a vector in Hilbert space, a complex-valued multi-dimensional vector space.

    So, if those dimensions are just accounting, doesn’t that require the SE to be epistemic — a description of something? Conversely, if the SE is ontological, then what is the physicality of a complex-valued multi-dimensional vector in Hilbert space?

    It seems that MWI does not associate itself with a Tegmarkian view, but it seems to give Platonic weight to the SE. But if the SE is not an epistemic description, but an ontological object, what exactly is it?

    (I’m always cautious and aware of my limits. FWIW, I may not be able to do the math or define a wave-equation on my own, yet, but I have gotten to the point I’m following along and getting teacher questions right. I’ve stepped back now to focus on improving my math so I can try a second pass with a deeper understanding of the actual mathematics.)

    ((I don’t know if this matters, but, had I not gotten into the arts in high school and software design in college, I was headed for being a scientist (since I was a young kid). I took German in high school for the express purpose of being able to read scientific papers in the original German. I ended up going down the software path, but I’ve kept my eye on those other ones. 😉 ))

  • Wyrd Smythe

    This final thread is for superposition and spin and whatever else remains…

    “Superposition of waves is a normal thing. Superposition of things like spin strike me as very strange, but spin already seems like a very strange thing anyway.”

    Indeed! There is an interesting consequence of matter waves and our understanding of them, though.

    Consider the two-slit experiment with water waves. The interference pattern is visible in the water itself, as well as along the “detector screen” (presumably a wall the waves impact). In the physical medium of the water, nodes (points where the water level is “zero”) are the result of cancellation of wave energy — a crest and trough combine and cancel.

    Importantly, the nodes we see along the wall (“detector screen”) represent this cancellation of energy. There’s no lack of water there, obviously, and there’s also no lack of wave.

    But in a quantum two-slit experiment, particles never (or rarely) land in the nodes. What’s zero there is the probability of a photon landing. The waves mean different things between classical and quantum theories — classical waves are physical transfers of energy; quantum waves, as far as we can tell, are probabilities.

    FWIW, it’s tempting to think of the two-slit as working with light waves exactly as it works with water waves — to even imagine the frequency of the light waves is what’s interfering. That isn’t the case. A free photon in flight has a well-known momentum (and thus an unknown position), and that momentum is described (in the SE) as a traveling plane wave. It’s this momentum description that interferes.

    Spin is especially strange (besides what it physically is) in that any spin state can be seen as a superposition of two opposing orthogonal states. A spin-up particle is also a superposed spin-left plus spin-right particle. This is true for any axis on which spin has been measured — measuring on the orthogonal axis has 50/50 probability.

    This behavior is actually true of all quantum conjugate pairs. As just mentioned, a photon in flight has a specific momentum, but an unknown position. Localizing the photon means the momentum value becomes unknown.

    “A lot hinges on whether we ever find evidence for an actual cut off point.”

    Exactly so. Perhaps cut-off condition is a better word. Everett allowed for several possibilities that would invalidate his view, one of which was such a cut-off point. Everett cast it as a magical N that, obviously, no one was finding, so he discarded the possibility.

    I believe it’s more complicated — some set of conditions necessary to strip away quantum behavior. Quantum states do seem very fragile, so the question is what exact mechanism accounts for that. I think the basic idea behind objective collapse theories is going in the right direction.

    I haven’t gotten to that Quanta article, yet.

  • Wyrd Smythe

    There is, perhaps, a fifth topic. You’ve questioned why physicists, especially those who don’t like the MWI, haven’t raised these issues, and I wonder that, too.

    The implication, obviously, is that I must be missing something, but nothing I learn ever seems to point to it. Sean Carroll wrote a whole book promoting the MWI, but it doesn’t answer key questions. (And doesn’t even address others.) My fascination comes, in part, from trying to figure out the sociology of it.

    I keep thinking about how everyone, including experts, once “knew” that ‘Eskimos have 50 words for snow.’ Except they don’t, and it was a mistake on one expert’s part that everyone else copied. They just assumed he was right.

    I wonder if something similar happens with MWI (and other fantastic speculation) and if there isn’t also a component of, as Sabine Hossenfelder put it, being “lost in math.” MWI, in particular, seems fixated on the Schrödinger equation — that fixation being promoted as a positive feature of the interpretation.

    So I dunno. I could be completely full of shit, but I might be also one of the few clear thinking people on the topic. I’m certainly not the only one who doesn’t agree with the MWI.

    Time, ultimately, will be the judge of that. I’m going to have to get on that Ask a Physicist page and ask some questions…

  • SelfAwarePatterns

    Sorry Wyrd. My ability to deal with thread sprawl is limited at the best of times. This weekend, the best I’ll be able to do is respond selectively and somewhat scattershot.

    On entanglement, consider if we have particle A is a superposition of spin up and spin down, so Aup + Adown. Now, we have it interact with particle B. In most interpretations, no collapse happens. (A collapse does happen in RQM, but only relative to each particle.) So now the entangled pair can be described by a common wave function: AupBup + AdownBdown + AupBdown + AdownBup. If we add in a third particle, it also joins the collection. Each particle has a state corresponding to each state of the other particles in the overall entanglement.

    Under collapse interpretations, as soon as these particles come into contact with a sufficiently sized or complex system, they collapse into a definite classical state. Under MWI, this never happens. The particles in the measurement device (screen, detector, etc) are no different than the particles in the system in question, so they join in on the entanglement. The original particle, A, is now entangled with the measuring device. This continues spreading into the environment.

    The thing about an absorbed photon is it always has an effect, causing an electron to move up into a higher energy orbital, which often causes it to emit another photon to move back down, or if imparted with enough energy, for it to escape its nucleus entirely. So the effects of the photon don’t just end if it’s absorbed.

    If only one particle is sent through the two slit experiment, it does interfere with itself, right.

    The case of photons interfering across branches seems more like Deutsch’s view, which I’m not currently onboard with.

    I think every interaction generates a new branch, but not necessarily one macroscopically different. Most interactions cancel each other out. So lots of macroscopically identical branches with one microscopic difference. Macroscopic differences happen when something (like a measuring device) magnifies the results of an individual or small number of interactions.

    The superposition can expand at up to light speed. It can also be contained under isolated conditions (like in a quantum computer). But for locality and the EPR thing, the best answer I can give is that the compatible results are coherent with each other and so interact when they meet. The incompatible ones aren’t, and so don’t. With spacetime branching, we could also say the compatible combos are in the same spacetime branch and the incompatible ones in others.

    On the normalization, I think it works out if you think it through. Sorry, I don’t have the energy to work through each one.

    I do think it’s a good idea for you to hit that Ask a Physicist site. Although you may want to prioritize. He may not respond if you hit him with a blizzard of stuff.

    • Wyrd Smythe

      This is good; we’ve locked down a few things…

      We agree the matter wave of a particle in flight interferes with itself, and that if branching occurs per the MWI every quantum interaction results in a branch (even if most of them are indistinguishable except for that interaction). The critical point is that the particle’s matter wave interferes with itself.

      As an aside, what did you mean by, “Most interactions cancel each other out.” Cancel in what sense? Are you just referring to their similarity or something deeper?

      “On entanglement,…”

      I’m not sure I follow these two paragraphs. I’m not sure what they are responding to. I do understand what MWI asserts. What I’m hoping here is to dissect it and examine its assertions.

      “So the effects of the photon don’t just end if it’s absorbed.”

      Of course not! Energy is transferred, at the least. What happens next, as you indicate, depends on the nature of the substrate. The photon may eject the electron, cause it to go wandering in the conductive substrate, or jump to a higher energy level and then back down releasing a new photon. In photographic film, the electron wanders, is captured by, and reduces an atom silver halide. (I think in some cases, you can get two photons each with half the energy.)

      What does necessarily end is the SE describing the photon. The energy of the photon system is transferred to the electron’s wave function. Other quantum information, may be transferred as well, although I’m not sure to what extent.

      “The superposition can expand at up to light speed.”

      The use of “can” there raises a question. Is it like a light cone, which always expands at light speed, or is it only passed through some form of communication (and therefore potentially slower)? What exactly is it that expands?

      What makes the spread of Alex’s |1⟩ measurement compatible with Blair’s spreading |0⟩ measurement, but incompatible with Blair’s |1⟩ measurement? Per the two-slit, we agree a matter wave interferes with itself, but this is different. What is the physics?

      I recognize you don’t have an answer. I also know you don’t like non-locality, but MWI doesn’t seem to require it.

      There is also that, if the SE is the central tenant, do you dispute that fully entangled particles are described by the same wave-function? How can a single quantum state described by a single wave-function be different in two places?

      “On the normalization, I think it works out if you think it through. Sorry, I don’t have the energy to work through each one.”

      I’m not asking you to; I’m asking you to hear and think about what I’m saying.

      The logic chain is simple: The proposition is that energy is thinned out but normalized. E=mc2 directly relates energy and mass, therefore mass must be thinned out and normalized. But gravity depends on mass. Free-fall rate depends on gravity, therefore free-fall rate depends on mass. Free-fall is measured in terms of distance and time, so if gravity changed, shouldn’t free-fall rate change?

      • SelfAwarePatterns

        On interactions cancelling each other out, I just mean that they don’t affect the macroscopic aspects of the branch. In most cases, the branches only differ by those microscopic differences.

        The paragraphs on entanglement were in response to this request above:
        “Can I ask you to elaborate on what you mean by “becomes entangled” with the (first) screen and “spreads throughout” the screen and environment? What, exactly, entangles and spreads?”

        On the spread of the superpositions, its spread depends on its ability to interact. If held in isolation, as we do with quantum computers, it doesn’t spread. So “can” was a deliberate word choice. In a Wigner’s friend experiment, it would be constrained to the isolated lab until it was opened and allowed to interact with the environment.

        On the locality stuff, it seems like we previously discussed this into the ground. At this point, I can only suggest checking out those papers I referenced before.

        I did note above that entangled particles can be described by a single wave function.

        “How can a single quantum state described by a single wave-function be different in two places?”

        Sorry, I don’t understand this question.

      • Wyrd Smythe

        Okay, I think I have a good picture of your view. All quantum interactions cause superposition even if most are effectively indistinguishable and superposition spreads strictly by interaction. It’s basically what I understand as Everett’s view.

        I think we agree it supervenes on the ability of classical objects to be in superposition. If that’s ever falsified, the MWI is falsified.

        Locality is an additional aspect, MWI doesn’t require it. What you saw as discussing “into the ground” I saw as taking it to about this point — establishing the view. The question is whether you’re willing to dissect it and examine possible mechanisms behind it. We’ve only ever gotten as far as the claim that it happens. I’d like to look closer; are you willing?

        “I did note above that entangled particles can be described by a single wave function.”

        With fully entangled particles, such as used in Bell’s Inequality experiments, it’s not “can” but “are” described by. The point of such experiments is that both particles share the same wave-function.

        So the question, “How can a single quantum state described by a single wave-function be different in two places?” asks about such a situation. Alex and Blair share a wave-function in each having a fully entangled particle. What these experiments appear to demonstrate is that if Alex interacts with the wave-function (thus changing it), Blair sees that change instantly.

        Bell’s Inequality experiments are valid in both single- and multiple-worlds interpretations, so locality is an addition — one I’d like to dissect if you’re willing.

      • SelfAwarePatterns

        My mind remains clouded, so it may be a day or two before I can engage in careful thinking about different scenarios.

        But I think the thing to always remember is that the MWI is just QM without the collapse. No collapse, no non-locality. You might stump me with a scenario, but that would just be me not knowing how the formalism works in that scenario. (And not knowing the math, I can’t fall back on it to calculate through it.)

        “What these experiments appear to demonstrate is that if Alex interacts with the wave-function (thus changing it), Blair sees that change instantly.”

        My understanding is that in collapse interpretations, the only interaction Alex can have with the wave-function that would affect Blair’s portion is a measurement. And Blair will only see the results if they do their own measurement. (I think anything else would violate the no-communication theorem.)

        In MWI, again in my understanding, the entanglement amounts to correlations set at the initial interaction. (All combinations of the correlations in superposition.) Since there’s no collapse, the entanglements never go away, but decoherence does spread them around.

        From what I’ve read, Bell’s theorem implicitly assumes single outcomes of measurements, which don’t apply to the MWI. For a discussion on how this works out, similar to the one I laid out in our earlier discussion, check out:

      • Wyrd Smythe

        I don’t know if you remember, but not long ago we talked about measurement as an interaction that affects the wave-function of both the measured and measuring system. This is true even in the MWI where the two become entangled. The measurement causes an abrupt change to the wave-function at the measurement point. In the MWI, this is a branch point.

        So MWI does have a form of “collapse” — a spacetime event where the wave-function suddenly changes. For instance, a system with an unknown spin (thus in superposition of all possible spins), upon measuring the vertical axis, branches into |U⟩+|D⟩ — repeating the same tests in the respective branches necessarily returns the same respective result with no further branches due to the repeated identical measurement(s).

        In either branch, testing a different axis does cause a branch, because |U⟩ or |D⟩ is a superposition of spins on other axes. As the hedweb link mentions, measuring orthogonal axes gives a random result. (Non-orthogonal axes have a correlated result — the closer to the vertical, the more likely to get the same |U⟩ or |D⟩ result as previously.)

        “My understanding is that in collapse interpretations, the only interaction Alex can have with the wave-function that would affect Blair’s portion is a measurement. And Blair will only see the results if they do their own measurement. (I think anything else would violate the no-communication theorem.)”

        I sorry, I don’t follow. The only interaction possible for Alex with the wave-function is a measurement, regardless of interpretation. Likewise Blair. All scenarios, even the hedweb link ones, involve Alex and Blair interacting with an entangled particle system.

        The form of non-locality involved doesn’t violate the no-comm theorem, since there’s no way to use this to communicate. Even when Alex and Blair agree on axes, they get mutually random results. It’s only when the compare them that they see a correlation. In the case of measuring the same axis, the correlation turns out to be 100%.

        “In MWI, again in my understanding, the entanglement amounts to correlations set at the initial interaction.”

        I believe that’s the case in all interpretations. I haven’t looked into how they produce spin-entangled electrons, but in photon experiments they use a down converter that splits one photon into two, each with half the frequency. Since they were originally one photon, the pair are fully entangled, at least until they interact with something else.

        (One thing that happens in single-photon experiments is that it gets lost because it hits a gas molecule or happens to interact with the walls or whatever. Another problem, apparently, is what they call “dark detection” — false detection due to noise and Heisenberg. When looking for a single photon, they use a clever trick: down convert it to two and require detecting both (called coincidence detection). They ignore single detection from either detector. Cool idea!)

        “From what I’ve read, Bell’s theorem implicitly assumes single outcomes of measurements,”

        I’m not entirely sure what to make of the single-outcome objection. As I see it, the predictions of QM — including non-locality — work out fine in the MWI. Locality is an additional requirement.

        I think for brevity and separation I’ll address the hedweb link in a second reply.

      • SelfAwarePatterns

        I do remember your stance that there’s a collapse under MWI. Sorry, whatever you have in mind, I don’t think it’s actually the MWI. (At least not the standard variants.) The closest thing to a collapse would be an observer on a particular branch losing access to portions of the wave (such as the other spin direction) through decoherence. But while decoherence is extremely fast, it’s not the instantaneous non-local collapse of Copenhagen, GRW, etc.

        “The only interaction possible for Alex with the wave-function is a measurement, regardless of interpretation.”

        But can’t a quantum system be manipulated without doing a measurement? Isn’t it how entangled particles are created in the first place? And of course, isn’t that what quantum computing is all about?

        But I was responding to your statement above that made it sound like Alice could manipulate her part of the wave function in a manner in which Blair could notice. Based on your response here, it sounds like I misinterpreted what you wrote. Sorry, my bad.

        For entanglement in collapse interpretations, my understanding is the correlations aren’t set until the collapse. That’s the non-local part. It’s what EPR argued against but the Bell inequality experiments showed to be the case. MWI only gets away with being correlated from the beginning because the combinations are spread over multiple branches. (I’m not sure how RQM handles this. Supposedly it’s local too.)

      • Wyrd Smythe

        “But while decoherence is extremely fast, it’s not the instantaneous non-local collapse of Copenhagen, GRW, etc.”

        Doesn’t that apply to any measurement, MWI or not?

        What difference exists for Alex upon making a measurement? In both the MWI and the CI, The wave-function changes very rapidly. (As far as anyone can tell, instantly.) The only difference I see is that, in the MWI, there are two copies of Alex both seeing a “collapsed” wave-function.

        Why can’t measurement in the CI avail itself to the same decoherence mechanism?

        “But can’t a quantum system be manipulated without doing a measurement?”

        Well, sure, but that doesn’t apply to the scenarios we’re discussing. As you go on to say, it was a misinterpretation. (I’m hurt that, after all you’ve read from me about quantum, you’d could think I’d mean that. I’ll blame it on your painkillers! 😀 )

        To be clear, the non-locality observed in experiments, even in theory, does not allow FTL communication. As an aside, virtual particles are also not restricted to locality. Virtual particle interactions can be FTL, but, again, it’s of no use in signaling.

        “For entanglement in collapse interpretations, my understanding is the correlations aren’t set until the collapse.”

        I’m not entirely sure I follow. Entangled particles are correlated because they’re entangled. Measurements demonstrate that correlation. (Could you be thinking of how Bell’s Inequality experiments use delayed choice in how the particle is measured? It’s part of closing various hidden variable loopholes.)

        It is the case that if both Alex and Blair measure the same axis — whatever angle that is — they get 100% correlated results. (Because entangled particles are described by one wave-function.)

        Likewise, if they measure orthogonal axes (again, any angle so long as they’re 90° separated), the results have 0% correlation. As I noted in the other reply, both these measurements — 0° and 90° — have results predicted by both classical and quantum theories.

        The quantum difference kicks in when the axes aren’t identical or orthogonal, but some other angle. Then classical and quantum theories give different predictions.

        It is, as you say, the non-local part. Entangled particles are described by a single wave-function. It’s exactly what Einstein hated but experiment confirmed. Importantly, note that this non-local behavior and tests of Bell’s Inequality are different things! As noted above, the correlation exists even in cases where the Inequality does not.

      • Wyrd Smythe

        I don’t know if this helps, but every theory I know forbids FTL communication of any kind. As you say, the no-communications theorem. (Special Relativity; Causality; FTL: pick two.)

        Saying MWI is non-local wrt to entangled particles still means knowledge of superposition is limited to light speed. In this case, knowledge of the results spreads at light speed from both Alex and Blair, but per the hedweb example, there are only two branches and in both the results are already correctly correlated.

      • SelfAwarePatterns

        I know you have a personal hunch otherwise, but the standard model of decoherence is not the collapse. They’re two separate things. Decoherence accounts for the disappearance of the interference effects. It provides no accounting of the other states ceasing to exist. Crucially, decoherence is a localized process. The collapse isn’t.

        I have read that there are versions of Copenhagen that incorporate decoherence. Can’t say I know much about them. But if those versions are not going to be MWI under a different name, they still have to have the collapse, the reduction of multiple states into one.

        On not allowing FTL, the extremely special nature of quantum non-locality actually makes it more suspect to me. If this was something we had converging evidence from other theories, my credence in it would be higher.

        I haven’t heard anything about virtual particles being non-local (other than through standard collapse non-locality). Can you elaborate? Or provide a link or reference?

        On entanglement and correlation, okay, I did conflate two things. One is the correlation in the wave function of all the possible combinations. That does happen in every interpretation. The other was the correlation in measurement outcome. It’s the latter which isn’t supposed to be set until the collapse in Copenhagen, and is what all the fuss is about. In MWI, the outcomes just become decohered from each other.

        It’s worth noting that there are two types of non-locality that often come up in quantum physics. One is action at a distance. It seems widely acknowledged that MWI does not to have this form of non-locality. The other is inseparability, the fact that a full accounting of a state requires knowledge of disparate components. MWI does have this form of non-locality. Personally, I think calling this “non-local” adds confusion since it’s really just correlation across distances, and it’s not the real issue, but physicists, including some MWI advocates, do discuss it.

      • Wyrd Smythe

        I was holding off until you had a chance to (hopefully) read the response to the hedweb link, but I think this might help clear up a point of confusion:

        “It’s worth noting that there are two types of non-locality that often come up in quantum physics. One is action at a distance. […] The other is inseparability,…”

        I have always been speaking of the latter, and it’s the latter that Bell’s tests involve. I’m not even sure what the former kind is — can you give me an example of what you mean by “action at a distance”?

        The simplest statement is in the Wiki for quantum entanglement. (In particular, the sentence: “If a state is inseparable, it is called an ‘entangled state’.”)

        Entangled states is what I’ve been talking about. As you say, MWI has it. The hedweb link demonstrates it.

      • Wyrd Smythe

        In reply to your earlier reply:

        “I know you have a personal hunch otherwise, but the standard model of decoherence is not the collapse.”

        The CI doesn’t have any formulation for measurement collapse, so decoherence certainly isn’t a part of it. As you go on to say, there are versions of the CI that do incorporate it, so it’s not as if the idea is solely mine or completely out of left field. And note my take on it is more nuanced (see below).

        I had asked about Alex and what happens as a consequence of measurement. Essentially, what’s the difference? The notion that decoherence accounts for why branches are inaccessible is a theory that’s been added to Everett’s construction — decoherence isn’t a part of Everett’s formulation.

        I think we need to recognize that decoherence isn’t yet well understood and is a topic of research under all interpretations. As I believe you know, one of my key questions about the MWI is how matter is able to coincide — that seems nonphysical to me. The MWI answers “decoherence” but doesn’t yet have a real physical theory to explain this bizarre fact.

        So some caution is warranted in all interpretations regarding blanket statements about decoherence.

        More precisely, my question is: Alex(MWI) makes a measurement, branches, and very nearly instantly both versions of Alex(MWI) find themselves with an apparently “collapsed” wave-function due to some presumed theory of decoherence. Alex(CI) makes a measurement and equally instantly has a “collapsed” wave-function due to… something. If the two situations appear so similar, and if the MWI one is due to decoherence, why might not a similar mechanism be at play in a version of CI?

        FWIW: As I’ve said before, I think Everett was essentially right on the small scale. I think measuring a quantum system does cause what we might call a superposition at the precise point of the interaction. For a very short time, and for a very short scale, that “superposition” might exist and spread (at light speed) for several molecules. But the overwhelming number of essentially random quantum states of the detector very quickly swamps out this quantum behavior. Before the wave of “superposition” can spread very far, it has encountered and mixed with millions of quantum states.

        The MWI notion is that reality splits and the split expands forever and decoherence explains why there’s no connection between branches. I’m suggesting reality splits for a very brief instant, but decoherence swamps it out almost instantly.

        “On not allowing FTL, the extremely special nature of quantum non-locality actually makes it more suspect to me.”

        Because? I do understand that your intuition demands locality, but I really do think this is one time you have to reconsider your intuition. Quantum theory is pretty clear on the matter, and experiments have apparently confirmed the theory. (OTOH, Einstein never did so you’re in good company if you don’t. 🙂 )

        Let me emphasize again that nothing in MWI insists on locality. (And, as I explained, that hedweb link fails to illustrate it.)

        I’ll look for a reference to non-locality in virtual particles. It’s something that was mentioned in connection with Feynman diagrams probably in a book I read.

        “On entanglement and correlation, okay, I did conflate two things. One is the correlation in the wave function of all the possible combinations. That does happen in every interpretation. The other was the correlation in measurement outcome.”

        The two are the same thing. The reason measurements are correlated is because of entanglement — there is no “other”. It all possible correlations being in the wave-function that accounts for the correlated measurements.

        “It’s the latter which isn’t supposed to be set until the collapse in Copenhagen, and is what all the fuss is about. In MWI, the outcomes just become decohered from each other.”

        I think you might be misunderstanding those experiments. There is no latter. The correlation was always there due to the entanglement. That correlation just becomes apparent due to measurements.

        Are you thinking about how Alex’s measurement determines Blair’s (or vice versa). Because entangled pairs share a wave-function, any measurement affects both particles. Once measured, further measurements reflect the change.

      • Wyrd Smythe

        Re the hedweb link. This got long and detailed, but I hope you’ll take the time to read it and consider it.

        An initial observation: The text assumes MWI (and so will I). It also assumes locality, but as it turns out, it doesn’t illustrate it. That said, I’m delighted to have a detailed scenario to discuss (and really do hope you’ll read this analysis carefully).

        No issues with his breakdown of EPR or the basics of spin. The first thing I noticed is the figure immediately following the line, “To establish familiarity with the notation let’s take the state of the initial wavefunction as:”

        To illustrate the notation, fine, but if it’s meant to represent an initial state of the entangled electrons, it’s an illegal state. As he says a bit above, “spin operators in different directions form non-commuting observables.” Exactly. It’s not possible for a single quantum state to have both left and up eigenstates.

        Since he does this in all three examples, I think he really is saying the entangled pair have a definite (in this case illegal) state. Each example treats the initial systems as if they had definite states. But if they did, then outcomes would be determined (and they’re not, as the examples show).

        The first scenario, right after seems almost to ignore the MWI. His third example looks at the same scenario — respective measurements on orthogonal axes — and does describe what happens under the MWI. Maybe this first scenario is intended only for illustration purposes?

        In any event, I think the more correct scenario (assuming MWI) is:

        |Ψ⟩_1 = |Alex,*,*,Blair⟩

        |Ψ⟩_1 ⇒ {local-observation} ⇒ |Ψ⟩_2

        Where |Ψ⟩_2 =

        Which is what he gets into in the final example. (For simplicity, I’m ignoring the normalization math. Just assume equal probabilities.)

        FWIW, if we do assume the electron pair is in a known state, it can only be an |U⟩+|D⟩ combo along some previously determined axis. But |L⟩+|U⟩ — definite states on different axes — is not a legal state for an entangled pair.

        Note that in the four branches that result from the two measurements, the electrons are no longer entangled with each other (they’ve become entangled with Alex and Blair). Further measurements show no correlation. The four electrons in the four branches can have |L⟩+|U⟩ combos, but not as entangled particles.


        The second example considers measurements along the same axis. He shows this as the superposition:

        |Ψ⟩_1 = |Alex,(L+R),(R+L),Blair⟩

        Which is fine, but note the implied correlation between the particles. He expresses that also as:

        |Ψ⟩_1 = |Alex,L,R,Blair⟩ + |Alex,R,L,Blair⟩

        Which is also fine, but he seems to assume a definite quantum state. Assuming an unknown initial state, it is also true that:

        |Ψ⟩_1 = |Alex,U,D,Blair⟩ + |Alex,D,U,Blair⟩

        Along every possible axis. As such |Ψ⟩_1 is a superposition of all possible measurements Alex and Blair could make. So some caution is warranted in these initial states. As I said above, a better way to notate the initial state is:

        |Ψ⟩_1 = |Alex,*,*,Blair⟩

        After Alex makes a measurement he has:

        |Ψ⟩_2 = |Alex[L],L⟩+|Blair,R⟩ + |Alex[R],R⟩+|Blair,L⟩

        The square brackets enclosing the recorded measurement. His notation makes it clear the wave-function has branched into two and that Blair measuring on the vertical axis can only get determined results.

        [Note this is exactly what I mean in saying MWI has a form of collapse. |Ψ⟩_1 could result in an infinite number of outcomes. After Alex measures, |Ψ⟩_2 has a different set of possible outcomes.]

        Then Blair measures and gets the only possible results:

        |Ψ⟩_3 = |Alex[L],L⟩+|Blair[R],R⟩ + |Alex[R],R⟩+|Blair[L],L⟩

        Note the wave-function has not branched — there are still two branches. The measurement Alex made altered the wave-function such that Blair’s measurement on the same axis is determined. As these examples show.

        The final step, communicating the results, likewise, does nothing to change the wave-function. |Ψ⟩_4 is the same as |Ψ⟩_3, except that Alex and Blair have communicated their results and, in each branch, confirmed the correlation.

        One key point is there are never more than two branches. The other is that communicating the results has no effect on those branches. The separation occurred when Alex made a measurement. |Ψ⟩_2 shows the wave-function is split, and |Ψ⟩_2 is indeed the correct description of the wave-function after Alex’s measurement. (But |Ψ⟩_1 is just one of myriad possible superpositions.)

        His example actually demonstrates non-locality.


        The third example considers orthogonal measurements. In this case quantum theory (in any interpretation) says the results aren’t correlated at all.

        He makes the same implication that |Ψ⟩_1 is definite rather than unknown. The superposition he specifies is just one of myriad. (Spin can be measured at any angle, so there exists such a superposition for every possible angle.)

        However |Ψ⟩_2 is definite because the wave-function has interacted with Alex and Blair. His description of |Ψ⟩_2 is correct — there are now four branches. (And four lights! 😀 )

        Importantly, after measurement, there are four branches throughout the description. He’s trying to prove locality and demonstrating the opposite.

        The basic problem is the assumption the initial wave-function is definite. It’s definitely not!


        The second example, matching axes, is the key of the three. That’s where correlation is 100% and non-local behavior is most apparent. When measurements are orthogonal the change to the wave-function isn’t apparent at all.

        Measurements on different but non-orthogonal axes are correlated, but not 100%. Bell’s Inequality deals with such partially correlated measurements because that’s the domain where classical and hidden variable theories give different predictions from quantum theory. (On matching and orthogonal axes they predict the same thing.)

        In that second example, the critical point is, from |Ψ⟩_2 on — once Alex makes a measurement — there are just two branches. Nothing later changes that. This is exactly what MWI (as a non-local) theory predicts.

        The proposition from you I understood went something like this:

        |Ψ⟩_1 = |Alex,*,*,Blair⟩

        Alex measures:

        |Ψ⟩_2a = |Alex[U],U,*,Blair⟩ + |Alex[D],D,*,Blair⟩

        Blair measures:

        |Ψ⟩_2b = |Alex,*,U,Blair[U]⟩ + |Alex,*,D,Blair[D]⟩

        In both cases, due to locality, the other’s measurement is unknown.

        Things now get complicated if they try to communicate because there are two illegal combinations (I’ll take the particles themselves out of the equation since they’ve been recorded):

        |Ψ⟩_3a = |Alex[U],Blair[?]⟩ + |Alex[?],Blair[U]⟩
        |Ψ⟩_3b = |Alex[U],Blair[?]⟩ + |Alex[?],Blair[D]⟩
        |Ψ⟩_3c = |Alex[D],Blair[?]⟩ + |Alex[?],Blair[U]⟩
        |Ψ⟩_3d = |Alex[D],Blair[?]⟩ + |Alex[?],Blair[D]⟩

        The |Alex[U],Blair[U]⟩ and |Alex[D],Blair[D]⟩ states (3a and 3d) are illegal and can never meet. The other two (3b and 3c) are legal and can meet.

        It’s not the scenario described by the link, but is it a fair description of the one you presented?

        If so, the question I’ve been asking is: What mechanism prevents |Ψ⟩_3a and |Ψ⟩_3d but allows |Ψ⟩_3b and |Ψ⟩_3c?

      • SelfAwarePatterns

        Okay, sorry, I don’t have the energy for a blow by blow critique, so I’m just going to focus on your final question, which I think gets to the heart of the matter, and which I’ll summarize as, under the MWI, what prevents uncorrelated results? (If that’s not the question, let me know.) Of course, this will be according to my understanding, so take with that in mind.

        The problem is seeing the world splitting at each site as independent events. They’re not. But they also don’t influence each other in real time. How then are they related?

        Remember, when the entangled pair of particles are prepared, all the possible correlations are set and exist in the wave function. They are the branches. They’re just not decohered from each other yet. If Alex and Blair did a measurement right then and there, the branches would decohere and begin spreading.

        But instead the particles are separated, that is, the isolated system modeled by the wave function is fragmented and separated across a spacetime interval. But the branches are still there the whole time.

        If Alex then performs a measurement, the branches begin spreading at their site. When Blair performs their measurement, the same branches begin spreading at their site. But the branches weren’t created at these sites. They were created when the entanglement was prepared.

        Why don’t incompatible branches interact with each other? The same reason branches in general don’t interact with each other under the MWI: decoherence. Why do disparate parts of a branch interact with each other? Because they are coherent with each other. This isn’t a unique situation for distantly entangled particles. It always applies. It’s just more obvious in this case.

        Last time I explained this, you were incredulous that decoherence could do this. But this is what decoherence is. If you doubt it and want more details, the thing to do is dig into the details of decoherence. (I wish I had those details, but don’t.)

        But consider a quantum computing circuit of qubits. Each qubit entangled in the circuit has states that are correlated with each other qubit’s states. That’s necessary for the massive parallelism in quantum computing. I think decoherence is simply this same phenomena writ large across the environment, in this case an environment spanning light years.

      • Wyrd Smythe

        “Of course, this will be according to my understanding, so take with that in mind.”

        Yes, and to be horridly blunt I’m questioning your understanding. Note, for instance, the confusion regarding entanglement non-locality, which it now appears you agree does exist in MWI. As such I think this response is part of that confusion.

        “The problem is seeing the world splitting at each site as independent events. They’re not. But they also don’t influence each other in real time.”

        They are definitely not independent because they involve entangled particles (an inseparable quantum system). Every experimental indication is that they do influence each other in real time. If you say they don’t, you’re floating a theory without supporting evidence. You’re free to believe it, but current experimental evidence says it’s wrong.

        (I guess you really are like Einstein. You just can’t let go of locality.)

        “Remember, when the entangled pair of particles are prepared, all the possible correlations are set and exist in the wave function. They are the branches.”

        Yes, and also remember there are an infinite number of possible potential branches. As we agreed with regard to the two-slit experiment branching doesn’t occur until interaction. There are an infinite number of possible outcomes implied but not realized by the wave function.

        Only interaction realizes particular branches. And note that making such a measurement changes what branches are possible in the future.

        “But instead the particles are separated, that is, the isolated system modeled by the wave function is fragmented and separated across a spacetime interval.”

        Yes, but the entangled particles form an inseparable state. A single wave-function.

        “But the branches weren’t created at these sites.”

        But we’ve agree in the context of the two-slit experiment that branches are created by interactions.

        “Why don’t incompatible branches interact with each other?”

        Let’s be clear you are not describing the scenarios illustrated by the hedweb link. You are apparently confirming the scenario I presented last, and giving me the same one-word answer: “decoherence” (And it’s that answer I’m trying to dissect and examine.)

        You’re right, I’m incredulous. I want to see physics that explains how decoherence allows two different branches of physical reality to meet or never meet. Decoherence is just what the name says: the loss of coherent phase information and thus the loss of interference patterns.

        Note also that we’ve agreed matter waves interfere with themselves, so coherent interference in this regard is not related to the MWI. Note also that when two systems lose coherence they do so in a single world and still interact with each other — just not coherently.

  • SelfAwarePatterns

    “Yes, and to be horridly blunt I’m questioning your understanding.”

    That’s okay, since it’s your understanding I think is problematic. 🙂

    I will say that I know my understanding has limitations, but I think I have a good handle on what they are and when I’m just accepting what the experts say. I’m concerned that you don’t have an appreciation of yours.

    The point of the 1935 EPR paper was, per Einstein, “spooky action at a distance”. From what I’ve read, he wasn’t worked up about inseperability. And the point of the Bell inequalities is demonstrating that to violate them, short of many-worlds or superdeterminism, there has to be action at a distance.

    “ Every experimental indication is that they do influence each other in real time. ”

    That only happens during the collapse in a collapse interpretation. (Which is the mindset they’re usually reported under.)

    “(I guess you really are like Einstein. You just can’t let go of locality.)”

    I will if the evidence decisively goes that way. As it stands, the evidence only requires it under the collapse postulate. Maybe someone will find evidence for the collapse. If so, I’d have to accept non-locality.

    “As we agreed with regard to the two-slit experiment branching doesn’t occur until interaction.”
    “But we’ve agree in the context of the two-slit experiment that branches are created by interactions.”

    You stated that as an agreement. The closest I came was noting that the language is often that branching happens with decoherence. But that’s just a language convention. The post-decoherence “branch” is a continuation of the pre-decoherence superposition “states”.

    “Note also that when two systems lose coherence they do so in a single world and still interact with each other — just not coherently.”

    Are we talking about quantum decoherence? If so, what would be an example?

    Decoherence by name wasn’t part of Everett’s thesis, but from what I understand, he used / borrowed from Bohmian mechanics. The wave dynamics of that are usually described as an early version of decoherence. (Decoherence was actually initially developed to support pilot-wave. It just turned out to be very useful for MWI, since MWI is pilot-wave minus the particle.)

    “If the two situations appear so similar, and if the MWI one is due to decoherence, why might not a similar mechanism be at play in a version of CI?”

    Taking only standard decoherence, I think that would make it the MWI.

    On Everett possibly being right on the microscale but wrong on the macroscale, could be. I’ve said as much myself. But you have to add assumptions to make the macroscale effects go away.

    “Are you thinking about how Alex’s measurement determines Blair’s (or vice versa).”

    Right. That’s the action at a distance. Under Copenhagen and with respect to Bell, the outcomes aren’t set until the measurement. And it’s what does not happen under the MWI. (While the inseparability aspect remains.)

    • Wyrd Smythe

      “That’s okay, since it’s your understanding I think is problematic. :)”

      😀 😀 Could be. I do constantly test and validate myself against new learning. In that context I seem to have a good handle on what I understand and what I don’t. I’ll happily stack my understanding against anyone’s. Bring it on! 😉

      “The point of the 1935 EPR paper was, per Einstein, ‘spooky action at a distance’.”

      That was something Einstein hated in general about QM. I brought up the EPR paper here because it illustrates that the phenomenon behind non-locality was part of QM at that time. The paper isn’t actually about non-locality but about entanglement (the phenomenon).

      The idea was that, with two entangled particles, one could measure non-commuting (conjugate) pairs, such as momentum and position, and thus apparently have knowledge QM forbids. Distance wasn’t a factor in the paper — the measurements could be done side-by-side.

      The point was that, if this was true, QM couldn’t possibly be a complete theory. Here’s the concluding paragraph from the paper (emphasis mine):

      Previously we proved that either (1) the quantum-mechanical description of reality given by the wave function is not complete or (2) when the operators corresponding to two physical quantities do not commute the two quantities cannot have simultaneous reality. Starting then with the assumption that the wave function does give a complete description of the physical reality, we arrived at the conclusion that two physical quantities, with non-commuting operators, can have simultaneous reality. Thus the negation of (1) leads to the negation of the only other alternative (2). We are thus forced to conclude that the quantum-mechanical description of physical reality given by wave functions is not complete.

      As I say, I only brought it up to show that entangled systems described by a single wave-function are part of the raw quantum formalism.

      Bell’s Inequality, as mentioned previously, is about differentiating quantum behavior from classical behavior. The behavior of entangled (inseparable) quantum systems is distinct from that.

      “You stated that as an agreement.”

      Part of the dialectic is that, if I misstate your position, you’re expected to correct me. The first goal is to explore an idea and discover the points of agreement and disagreement. We have to be clear about our respective positions to accomplish that.

      I went back and looked at the conversation. In your first reply you said:

      |SAP⟩ I do think decoherence is a natural event to speak of a “split”, since it’s where the interference effects become undetectable.

      When I sought clarification you said you didn’t agree with the Deutsch formulation, which does involve existing branches. I sought clarification again, and you replied:

      |SAP⟩ I think every interaction generates a new branch, but not necessarily one macroscopically different.

      Which seems kinda definite. Branches are due to interaction. (That’s Everett.)

      I’ve repeatedly invited a specific account of what happens in the two-slit experiment. If I’ve misunderstood your position, at this point it’s kinda on you. I’ve tried to understand your answers.

      “Are we talking about quantum decoherence? If so, what would be an example?”

      Yes. Every classical interaction involves decohered systems. You don’t interfere with the chair you’re sitting on. Or anything else you interact with.

      “Decoherence by name wasn’t part of Everett’s thesis, but from what I understand, he used / borrowed from Bohmian mechanics.”

      Not in his main thesis, but he references Bohmian mechanics in the Supplementary Topics and Discussion sections later in the paper. You should read the paper yourself, but one salient part is probably the paragraph:

      Bohm considers ψ to be a real force field acting on a particle which always has a well-defined position and momentum (which are the hidden variables of this theory). […] Our main criticism of this view is on the grounds of simplicity — if one desires to hold the view that ψ is a real field then the associated particle is superfluous since, as we have endeavored to illustrate, the pure wave theory is itself satisfactory.

      Interestingly, in that part Everett talks about the same sort of spin measurements I’ve been discussing (see Stern-Gerlach experiment). Everett suggests that merging the beams would re-create the original states, which would only occur under MWI. I wonder if anyone has tried that. It kind of reminds me of that discussion about reversing the spin measurement.

      One thing about Bohmian mechanics — it’s explicitly non-local. (FWIW, the pilot wave theory wiki page only mentions “decoherence” once in a parenthetical statement.)

      “Right. That’s the action at a distance.”

      This continues to be the crux of disagreement. You apparently don’t accept that two entangled particles form an inseparable state and are described by a single wave-function, despite this wiki link that says otherwise. The key phrase is: “If a state is inseparable, it is called an ‘entangled state’.”

      Secondly, we apparently disagree about the nature of the quantum system prior to measurement. The disagreement specifically involves the definiteness of an unknown state. I think you (and the hedweb guy) assign too much definiteness to it, and I think it reflects a misunderstanding of the wave-function.

      No doubt you won’t believe what I say, but here goes. I begin with Ψ_1, which describes an electron with unknown spin. It is true that:

      Ψ_1 = |U⟩ + |D⟩

      Which presumes a measurement on the Y-axis. (Again, I’m skipping the normalization math. Assume equal probability in all cases.) It is also true that:

      Ψ_1 = |L⟩ + |R⟩

      Which presumes a measurement on the X-axis. (We’ll consider the Z-axis the direction of travel. Particles that move sub-light also have spin on this axis. Particles that move at light speed do not.)

      Since spin can be measured on any axis, and the the result is either + or on the measured axis (we’ll call them 0 and 1), it is true for any angle that:

      Ψ_1 = |0⟩ + |1⟩

      The problem with assigning any definiteness to this is that different descriptions are incompatible. For instance, it is not true that:

      Ψ_1 = |U⟩ + |L⟩


      Ψ_1 = |U⟩ + |D⟩ + |L⟩ + |R⟩

      Because there’s no way to measure two axes at once; no way to get one result or the other. Orthogonal spins are non-commuting. This is true of all possible angles — none of them commute with each other.

      So Ψ_1 is not definite (which is a fundamental statement of QM). There is nothing one can say about spin value until one picks an axis and makes a measurement.

      In the MWI context, it would imply an infinite number of potential branches that begin with the creation of the particle. All those branches except for two then vanish due to the measurement. Ultimately, a spin measurement results in only two branches.

      One can say those two branches were in Ψ_1 to begin with, but they were there with every other possible measurement — all of which are incompatible. The superposition that’s relevant is the one that gets measured. Until then, all one is saying is that one could measure on a given axis, and if one did, one would get |0⟩ and |1⟩ branches due to that measurement.

      With an entangled pair, we have an inseparable state:

      Ψ_2 = Alex|01⟩ + Blair|10⟩

      For any correlated pair of axes. That is, if Alex and Blair measure on the same axis, then either Alex gets |0⟩ while Blair gets |1⟩, or vice versa.

      In MWI (as noted in the hedweb text), this results in two branches containing both those outcomes. We end up with:

      Ψ_3 = (Alex+Blair|01⟩) + (Alex+Blair|10⟩)

      Which takes this to the third point of disagreement: Are there two branches or four?

      This has gotten long so I’ll leave off with that question. (I have a thought experiment I’ll post in a new thread.)

      • SelfAwarePatterns

        I’m going to respond to your scenario below, but just a point of clarification. I try to make my responses in these conversations thoughtful, but that takes…well…thought, and effort. As a practical matter, I don’t have the time or energy to do that with every single thing I disagree with in every comment. Often it’s because my response would just be a repeat of previous responses. Just as often, it’s a matter of prioritizing and consolidation. Occasionally I just overlook it.

        All of which is to say you shouldn’t take not responding as agreement. Feel free to follow up if it’s something you really want addressed, but the answer might amount to: “What I said before”.

      • Wyrd Smythe

        You’re of course free to ignore anything you want, but as I just said: “Part of the dialectic is that, if I misstate your position, you’re expected to correct me. The first goal is to explore an idea and discover the points of agreement and disagreement. We have to be clear about our respective positions to accomplish that.”

        Failing to do that is acting in bad faith. It’s one thing to ignore some point I made, but when I’m trying to create checkpoints in the discussion so it can be moved along, it’s vital you participate in that.

        I’m trying to discover exactly what your views are, but it often feels like I’m pulling teeth. You don’t seem to like details and specifics.

        [As an aside, that link you posted above about being open or closed… I observe that your stance here appears 100% closed. You seem to feel local MWI is the One True View, yet both are, at this point, nothing but presumptions.]

  • Wyrd Smythe

    I think I understand the view you’re presenting regarding locality, so I’d like to lay out the scenario I hoped to dissect with you. It uses the same Alex and Blair spin-entangled particles experiment.

    A key point is the particles are fully entangled — a nonseparable quantum state. It seems we still need to resolve this point.

    Let me start with a non-local MWI description. Then I’ll do the local version I understand from you.

    Alex and Blair plan to measure and record 6 spin-entangled particles in succession. They agree on a set of six measurement angles. [e.g. 0°, 75°, 22.5°, 90°, 0°, 45°] We’ll assume considerable (light years) separation between them. They retire to their labs, receive six entangled particles each, and record their results.

    In the CI, each gets a random six-digit string that will turn out to be the opposite from the apparently random string the other got. They’ll have to meet again to discover this, though.

    In the non-local MWI, the measurements result in 64 branches. In each branch, Alex and Blair have an apparently random bit string that will turn out to correlate if they meet and compare them.

    Since the wave-function is nonseparable, their measurements are necessarily correlated, so there can only ever be 64 branches. (Due to these measurements. There must be an infinite number of unrelated branches going on (like virtual particles) due to all the quantum interactions that don’t matter at the classical level.)

    Now Alex and Blair pass their bit strings to, respectively, Chris and Drew. Both copy the bit strings and pass them on to Ev, who is the only one to see both results. As such, only Ev will see any correlation. Everyone else just sees random bit strings.

    The 64-fold superposition spreads to Chris and Drew separately, but they were both “seeing” the same inseparable quantum event — determining the spin of one of the six particles. (Obviously there is no problem with two people with space-like separation receiving information from a single light cone. This is the same idea, except here the light cone has two sources, Alex and Blair.)

    Two points about this set up: All parties have significant space-like separation. It takes time for the superposition to spread to each (through comms or physical medium). Note the various methods of communication all have their own quantum properties — only recorded information passes between them, not quantum information.

    Importantly in this scenario, branches never have to “meet or not meet” — there are no forbidden states and no need for branches to be “coherent” or not.

    It does mean accepting the non-locality of nonseparable entangled states. Which does come from the raw formalism.

    As I understand the local version, the same set up results in not 64, but 128 branches, because the measurements Alex and Blair make are separate. (I’m not clear on why, other than you insist they can’t due to locality. Do you believe the wave-function becomes separable? What breaks the inseparability?)

    Among these 128 branches only “mated” pairs can ever meet. A branch with an Alex who got [101010] can only meet a branch with a Blair who got [010101]. It can never meet a branch with any other result. Since the results have been copied to various mediums, it’s hard to see how coherence can play any role in what branches meet. How does six bits of information control which branches meet?

    As the superposition spreads, there are 64 unrelated versions each of Chris and Drew (each joins half the 128). Here again, though, only matching pairs of Chris and Drew can ever meet.

    Lastly, 64 messages come from Chris, and 64 more come from Drew, and here is where branches must meet carefully. There can only ever be 64 versions of Ev, and those versions can only receive one particular message from each of the two sets of 64.

    In this case, branches need to merge from 128 to 64, and they need to merge in specific way.

    And, yes, I confess I am incredulous. What I’m asking for is the specific physical mechanism that accounts for this behavior. Just saying “coherence” doesn’t.

    • SelfAwarePatterns

      I’m a little nervous about there being six sets of entangled particles in this scenario. That may be introducing complexities I’m missing. So sorry if I didn’t adequately address that aspect of it.

      It seems like a lot of this discussion has gotten hung up on language semantics. So just to be clear, under the MWI, whenever we talk about 64 or 128 branches, by the time any comparison is possible, we have vast multitudes of branches which among them have 64 or 128 variants related to the original system in question. (And you’re right, there are far vaster multitudes where none of these scenarios ever took place.) So when I say “there are X branches”, I mean that are X scenarios among all the relevant branches.

      Another language point. A “branch” is simply a continuation of a “state” in the original superposition of the system being measured. The only difference is that a “branch” is a “state” that has become massively entangled with the environment. This is a crucial point for the explanation below.

      So, in your second “non-local” scenario, there should only be 64 branches. There remains inseparability, but no real time interaction, no action at a distance. The inseparability amounts to a set of correlations, correlations set because the entangled particles share a causal history going back to when they were initially entangled.

      Focusing on just one of the entangled pairs, there are two relevant states in the combined wave function from the beginning. When later measured at Alex’s site, the local portion of each state becomes entangled with Alex’s environment. When measured at Blair’s site, the local portion of each there becomes entangled with Blair’s environment. It’s not a matter of them “merging” later during the comparison, but that they’re already entangled.

      It’s also worth remembering that under the MWI, there are no pure classical interactions. All interactions are quantum, because the whole universe is quantum. So the comparison is a quantum event as much as any other.

      When considering the specific physical mechanism, there is nothing happening here that doesn’t happen in a circuit of qubits. In a 53 qubit circuit, the last circuit can be entangled with the first even with dozens of qubits in between. Executing algorithms in parallel wouldn’t work if that weren’t true. In the case of a branch interacting with itself, it’s the same thing, just with a lot more particles involved.

      • Wyrd Smythe

        “I’m a little nervous about there being six sets of entangled particles in this scenario.”

        The purpose was to establish a “random” bit string and also to emphasize the problem of later combining branches. (You didn’t address that, but it was the main question.)

        “It seems like a lot of this discussion has gotten hung up on language semantics.”

        If so, it’s due to lack of detail and specificity. I’ve tried, but you need to participate fully in that process.

        “(And you’re right, there are far vaster multitudes where none of these scenarios ever took place.) So when I say ‘there are X branches’, I mean that are X scenarios among all the relevant branches.”

        Agreed, and that has always been the case. (If you’re referring to what I think you are, I didn’t mean branches where the scenarios never took place at all; I was referring to the myriad of indistinguishable quantum branches due to trivial quantum events that don’t change anything.)

        [BTW, one reason branching in MWI is a discussion point here is that, among the flavors of MWI, there are two major views depending on whether indistinguishable quantum events result in branches. Under one view, only amplified quantum events — ones that make macroscopic differences — result in branches. If I’ve understood you, you lean to the other view that says all quantum events create branches. It’s worth noting the latter view results in a truly mind-blowing number of branches. Every photon that lands somewhere results in a vast number. Every atom is spitting off branches at an unimaginable rate. It’s quite an ontological commitment.]

        “Another language point. A ‘branch’ is simply a continuation of a ‘state’ in the original superposition of the system being measured. “

        Here’s where we have a problem, because on my understanding of QM that’s just false. The system does not have a state until measured. It is a superposition of all possible states.

        And because most of those possible states are incompatible, it’s incorrect to assign any definite state to the system until it’s measured. Quantum computing, in fact, depends on this fact. The superposition includes all possible answers — it requires an interaction to select one.

        So we have a disagreement here. My understanding of QM is that an unknown quantum system does not have a definite state. It’s a superposition of any possible state resulting from some interaction.

        “So, in your second ‘non-local’ scenario, there should only be 64 branches. “

        Yes, as so stated, but you continue to deny the non-local nature of separability in what is explicitly the non-local scenario. (You again seem to believe the original system has a definite state. It doesn’t; it is only potential. The only initial condition is the fact of entanglement.)

        This is the second point of disagreement: You don’t agree with what separability (or entanglement) implies. (Which, as I’ve pointed out is what the QM formalism implies.)

        (Your desire for locality and initial conditions make it seem almost as if you want to fit a QM peg into a classical hole. It’s not a good fit.)

        “There remains inseparability, but no real time interaction, no action at a distance.”

        No. This scenario was explicitly non-local, so one measurement affects the entire quantum state instantly. That’s built-in to the scenario. (Again, it’s what the QM formalism says happens. Refer to any number of Wiki articles or to the Scott Aaronson document.)

        Here’s the problem: The non-local example is intended in contrast to the local example (which you ignored but it was what I understood your version to be). The non-local version is my non-local MWI version.

        Now either you believe in a 128-branch version as described in the local MWI scenario, or your version involves initial conditions such that there are only 64 branches. A key difference is that, in the 64-branch version, the branches remain separate (“decohered”). There are no issues with combining the right ones — none of them ever combine.

        But in the 128-branch version, which is strictly local, then there is a situation where only a branch from each sub-group (Alex vs Blair) can meet only one branch from the other sub-group. And these sub-groups do have to meet to inspect the results. (If Alex and Blair really are making separate choices that don’t affect each other, this necessarily results in 128 branches and requires merging them correctly.)

        So does the local MWI scenario not reflect your view? You now seem to describe what is essentially a hidden variables theory involving a definite state for an unknown quantum system. (That, BTW, is more aligned with the Deutsch view of pre-existing branches.) Please clarify.

      • SelfAwarePatterns

        Wyrd, when I wrote “in your second ‘non-local’ scenario, there should only be 64 branches”, I meant “local” rather than “non-local”. All of my remarks should be taken with that in mind. Sorry for the typo and confusion. (Not that I suspect you’ll be any happier with the result 🙂 ).

        “ The system does not have a state until measured. It is a superposition of all possible states.”

        These are Copenhagen tenets. In MWI, all the states in the wave function exist and are equally real. It’s the central idea of Everettian physics, QM without the collapse, the pure causal and continuous evolution of the wave function. As I noted above, this point is crucial. If you can’t accept it, I can’t see this or any other aspect of the MWI making any sense for you.

      • Wyrd Smythe

        Okay, so essentially a hidden variables version. Recalibrating…

        “So, in your {ed: local} scenario, there should only be 64 branches. There remains inseparability, but no real time interaction, no action at a distance.”

        But that is a contradiction in terms. Inseparability directly implies instant change (in any interpretation). How do you reconcile both inseparability and locality?

        Do you hold that, given sufficient time, the change Alex makes eventually alters Blair’s?

        An inseparable quantum state by definition is a single quantum state. How do you account for the electron pair Alex and Blair share being inseparable yet separable?

        “The inseparability amounts to a set of correlations,…”

        No, it doesn’t. I’ve pointed you to the definition of inseparability multiple times. That’s not it.

        We can disagree on this, but you’re also disagreeing with QM.

        “Focusing on just one of the entangled pairs, there are two relevant states in the combined wave function from the beginning.”

        We’ll also have to disagree on assigning any definiteness to the wave-function prior to interaction. You’re singling out the superposition you want, but it’s one of an infinite number.

        Note that, once an interaction has occurred and branches are real (under MWI), that’s different. But prior to the interaction, all possible branches exist — all possible correlated and non-correlated angles of measurement. Remember, all those superpositions are mutually incompatible!

        My question is, given that we can equally describe Ψ as |U⟩+|D⟩ or as |L⟩+|R⟩ (or as any angle |0⟩+|1⟩), when Alex+Blair measure the Y-axis and thus branch to Alex(|U⟩+|D⟩) + Blair(|D⟩+|U⟩) what happened to the |L⟩ and |R⟩ states (and the infinite number of others)?

        What makes Ψ=|U⟩+|D⟩ special? According to QM, it’s not.

        All I can say at this point is that local MWI seems like a double presumption to me. Experimentally, we see a non-local single world, so a belief that opposes both of those seems an extraordinary claim to me. I need extraordinary proof.

      • Wyrd Smythe

        Are you abandoning the discussion or researching your answer? As it stands, the wave-function point might be at least debatable, but treating inseparable states as separable is a logical contradiction. Are you denying inseparability?

      • SelfAwarePatterns

        From your previous reply, it didn’t sound like you were willing to budge on the wave function issue, and since that meant we weren’t talking about the same MWI, it seemed unproductive to continue butting heads over an aspect of it. (And honestly, the tone of the conversation was making me think we needed to take a break.)

        On inseparability, as I noted above when I described it, I don’t deny it. I do think you’re misinterpreting what it means. (Unfortunately, a lot of sources make this easy to do.) Here I’ll quote a few snippets from Wallace that I think encapsulate my understanding of it. (Sorry, should have done this earlier.)

        In fact (and here I largely follow Healey 1991; 1994) we can usefully distinguish two sorts of nonlocality, which are usually called action at a distance and nonseparability. Action at a distance occurs when, given two systems A and B which are separated in space, a disturbance to A causes an immediate change in the state of B, without any intervening dynamical process connecting A and B. (Conversely, theories which do not involve action at a distance are said to satisfy Local Action.)

        Wallace, David. The Emergent Multiverse (p. 293). OUP Oxford. Kindle Edition.

        Nonseparability is a matter, not of dynamics, but of ontology. A theory is nonseparable if, given two regions A and B, a complete specification of the states of A and B separately fails to fix the state of the combined system A + B. That is, there are additional facts—nonlocal facts, if we take A and B to be spatially separated—about the combined system, in addition to the facts about the two individual systems.

        Wallace, David. The Emergent Multiverse (p. 293). OUP Oxford. Kindle Edition.

        8.5.1 Does Everettian quantum mechanics display action at a distance? No. In a quantum field theory, the quantum state of any region depends only on the quantum state of some cross-section of the past light cone of that region. Disturbances cannot propagate into that light cone.

        Wallace, David. The Emergent Multiverse (p. 302). OUP Oxford. Kindle Edition.

        8.5.2 Does Everettian quantum mechanics display nonseparability? Yes. Because of entanglement, knowing the density operators of regions A and B does not suffice to fix the density operator of A ∪ B. Some of the properties of A ∪ B are genuinely nonlocal: they have local physical manifestations only if we arrange appropriate dynamics.

        Wallace, David. The Emergent Multiverse (p. 303). OUP Oxford. Kindle Edition.

        These largely agree with the other sources (such as Lev Vaidman) I’ve seen this distinction discussed.

      • Wyrd Smythe

        Yeah, we’re likely dead-ended on wave-function definiteness. We don’t seem even to agree the state of things is different (in all interpretations) before and after the interaction.

        I don’t know quite to make of the Wallace quotes. The first and second paragraphs are from the same page and seem describe inseparability. The first one seems to be where you got the idea that inseparability and action at a distance are distinct.

        When you mentioned this, I asked you what the difference was. I’ve been discussing inseparability — entanglement — all along. I have no idea what is meant by “action at a distance” if it doesn’t refer to inseparability. Every time I hear nonlocality discussed in physics, it’s inseparability that’s meant. So I’m still lost about what distinction is being made here or how that distinction applies.

        The second paragraph seems to continue the definition, but the language is strange. We’re not speaking about “A theory” we’re speaking about quantum mechanics, and as far as I know it’s the only theory in which inseparability exists. There are no other theories that have inseparable aspects — it’s a big part of what makes QM so weird. I also don’t understand the word “region” — we’re talking about quantum states, which I’ve never heard called regions. What does a “region” refer to?

        “…is nonseparable if […] a complete specification of the states of A and B separately fails to fix the state of the combined system A + B.”

        I don’t know what that means. In QM, inseparability=entanglement, and there is well-known math for it. I don’t understand this talk of “nonlocal facts” — you’ll have to explain.

        In paragraph three, “action at a distance” is something other than entanglement, so doesn’t directly apply to this discussion (as far as I know). I certainly agree about light cones with regard to causality.

        I can’t make head or tail of paragraph four, so please break it down for me.

        In any event, these seem more philosophical assertions than quantum physics. He’s saying things are true, but not demonstrating, let alone proving, them. And I have no idea what he’s saying is going on with inseparability.

      • SelfAwarePatterns

        The first two paragraphs are actually not the same definition. The first paragraph is just the beginning of the description of action at a distance. (It’s actually the end of one paragraph and the beginning of another, but the copy and paste from Kindle lost the formatting.) The action at a distance description goes on to mention things like de Broglie-Bohm theory and Newtonian gravity as examples.

        The second paragraph is much further down the page and begins a summary discussion on nonseparability. I think “region” just refers to a spacetime region.

        Copehagen has both action at a distance (collapsing to one “coordinated” state) and nonseparability. Since most of the discussions about QM nonlocality are within the context of Copenhagen, they usually don’t make the distinction.

        On the fourth quote, maybe these subsequent paragraphs will help. (Hopefully the formatting doesn’t get too wrecked.)

        For instance, suppose that we have a long row of qubits q1,… qn, and we simulate local interactions by only ever applying gates to adjacent pairs of qubits. The system might start in a state with no nonlocal properties: say, with each qubit in the state |00|. By applying a gate to the first two qubits, we can transform them so that their joint state is (the projector onto) any one of the four states
        {|X± = (|0⊗|0±|1⊗|1), |Y± = (|0⊗|1±|1⊗|0)}). (8.3)
        In each case, the states of qubit 1 and qubit 2 are each
        1 2 (|00|+|11|). (8.4)
        Now we can effectively transport qubit 2 along the chain by applying a sequence of swap operations between adjacent qubits. At the end of this local process, all qubits except q1 and qn are in the state |00|, q1 and qn individually are in the state (8.4), and q1 and qn jointly are in one of the four states (8.3). However, the joint state of q1 and qn is not locally accessible: no local operation (i.e. no operation on adjacent qubits) leads to local results (i.e. states of individual qubits) which depend on the joint state of q1 and qn.

        Finally, we can transport qubit 1 along the chain in the same manner. At the end of the process, qn−1 and qn will be in the entangled state, and this can be determined by a local interaction.

        Somewhat picturesquely, we can think of entanglement between states as a string connecting those states, representing the nonlocal relation between them. We can move either end of the string by local interactions, and we can cause the string to ‘fray’ at either end by entangling the system at each end with adjacent systems. But we cannot access the information content of the string—i.e. we cannot set up dynamical processes whose outcomes are dependent on the nonlocal properties represented by the string—without moving the two ends of the string until they coincide. In this way, nonseparability remains fully compatible with dynamical locality.

        Wallace, David. The Emergent Multiverse (p. 304). OUP Oxford. Kindle Edition.

        All of which leads me to think of pure nonseparability as more of an issue of accounting for correlated states. I’m open to the possibility I’m oversimplifying. But the key is it’s not action at a distance.

        On philosophical assertions, remember I’m giving you small snippets from a 32 page chapter, and out of a 530 page book, much of which, honestly, is above my head.

      • Wyrd Smythe

        Newtonian gravity is a good example of what I would call “action at a distance.” So is any form of FTL. Modern physics, let alone quantum physics, doesn’t have that — the no-comm theorem prevents it. As such, I would dispute that wave-function “collapse” in the CI is this. Mathematically, it’s a nonlinearity in a linear equation. Physically we don’t know what it is, but nothing is actually affected, nor can it be used to communicate.

        The CI has inseparable states, obviously. Wave-function collapse is essentially the same thing — a single wave-function describing a system changes so all aspects of what it describes must change at the same time. It’s not action at a distance in the usual sense — it never violates causality; action at a distance does. (If Newtonian gravity was true, we could communicate instantly with gravity waves.)

        I’m sure “region” does apply to a spacetime region, but what do spacetime regions have to do with quantum states? (I live in a region of the USA, and one can talk about a region of the local galactic group, so it’s a particularly vague word.)

        The part you quoted involves a scenario unlike what we’ve been discussing. His scenario involves quantum computing qubits with gates allowing actions on those qubits. If I follow what he’s saying, it’s something like this: If I start with an entangled pair of qubits and then perform a series of swap operations to move one N qubits away, then “the joint state of q1 and qn is not locally accessible.”

        However I’m not clear on what he means by that or by an operation on “adjacent qubits” (which adjacent qubits?) or what that has to do with Alex and Blair measuring an entangled state. What is the connection between quantum computing operations and the scenarios involving fully entangled pairs?

        Is he asserting that q1 and qN are fully entangled but interacting with either doesn’t affect the other? Doesn’t QC depend on that entanglement? And if both are described by an inseparable wave-function, how is it that interacting with one doesn’t affect the other? Again, I thought QC kinda depended on that sort of thing.

        “All of which leads me to think of pure nonseparability as more of an issue of accounting for correlated states.”

        I don’t think “pure” means anything with inseparable states — a quantum state either is separable or inseparable (entangled). The wiki page for Separable state is fairly brief and clear. The one-line lede is: “In quantum mechanics, separable quantum states are states without quantum entanglement.”

        The first section Separable pure states puts it in the nutshell, and the math isn’t too bad there. The key sentence is, “Otherwise it is called entangled. When a system is in an entangled pure state, it is not possible to assign states to its subsystems.”

        Every QM reference says the same thing: Inseparability is a fundamental aspect of the mathematics of QM. It’s built in to how the math works, and it’s been there since day one. It’s what Einstein hated because he thought it was action at a distance (but spooky).

        (I looked up the book. Wallace is a philosopher, which might explain why it felt that way to me. I can spot a philosopher a mile off. 😀 )

  • Wyrd Smythe

    I’ve been slowly working through Scott Aaronson’s Introduction to Quantum Information Science Lecture Notes. I believe you may have downloaded a copy, too? His section, 5.3.2 Entanglement, covers this entanglement topic.

    I’ll quote one part (he’s talking about a pair of entangled qbits — the situation is analogous to electrons with entangled spins):

    This state is particularly interesting because measuring the first qubit collapses the state of the second qubit. The state can’t be factored into a tensor product of the first qubit’s state and the second qubit’s state. Such a state is called entangled, which for pure states simply means: not decomposable into a tensor product.

    A state that’s not entangled is called unentangled or separable or a product state (for pure states, which are the only kind being discussed at this point, all three of these mean the same thing).

    The basic rules of quantum mechanics, which we saw earlier, force entanglement to exist. It was noticed quite early in the history of the field. It turns out that most states are entangled.

    As we mentioned earlier, entanglement was arguably what troubled Einstein the most about quantum mechanics. He thought that it meant that quantum mechanics must entail “spooky action at a distance.” That’s because, while typically particles need to be close to become entangled, once they’re entangled you can separate them to an arbitrary distance and they’ll stay entangled (assuming nothing else is done to them). This has actually been demonstrated experimentally for distances of up to 150 miles (improved to a couple thousand miles by Chinese satellite experiments, while this course was being taught!).

    Let’s say that Alice and Bob entangle a pair of particles by setting their state to 2^(-1/2)∙(|00⟩+|11⟩). Then Alice brings her particle to the moon while Bob stays on Earth. If Alice measures her particle, she can instantaneously know whether Bob will observe a |0⟩ or a |1⟩ when he measures his.

    He goes on to explain the quantum weirdness aspect of this rather nicely. (I’ll note this explanation exactly fits my understanding and is what I’ve been saying in this discussion.)

    Aaronson also has a section on the MWI, 12.4 The Many-Worlds Interpretation. His text seems (from what I’ve seen) interpretation neutral. The only place the word “MWI” appears is in that section.

  • Wyrd Smythe

    Say Mike: Do you favor the MWI because you want locality? I’m getting the impression you place a higher value on locality than on the MWI. If you had to pick one…

    • SelfAwarePatterns

      Actually, locality is just one of its attractions. Wave mechanics without the collapse postulate makes QM deterministic, local, and realist in a way that offers a cleaner account of quantum computing and is more broadly compatible with cosmology. (Obviously it’s not completely compatible since we still don’t have a theory of quantum gravity. But I suspect we’ll need a realist account of QM if we hope to get that.)

      For those only hung up just on locality, there are other options. But they’re not deterministic, keep the collapse postulate, along with all its weirdness, and have varying levels of anti-realness.

      Certainly the idea of not just particles, but the whole universe in superposition is disconcerting. But those broader implications are untestable. Focusing on testable predictions, I’m impressed that Everettian physics passes them with fewer assumptions.

      Doesn’t mean someone won’t discover an actual collapse tomorrow and falsify it. Indeed, as I’ve said before, if the Everett is right, I suspect he’s right in the way Copernicus was right in 1543, with a lot of additional paradigm shifts still necessary.

      • Wyrd Smythe

        “Actually, locality is just one of its attractions.”

        I think I noted you want to force QM into a classical hole, and apparently it really is the case that you want a classical universe. (Whereas I’m fine with, and rather like, quantum weirdness, but should I make it clear I’m not advocating the CI or saying it’s a complete answer? This is about issues I perceive in the MWI.)

        I’ve had an interesting morning (in a good way). You’ve asserted before that the MWI is favored in QC and cosmology, but I’ve don’t see it mentioned in mainstream books or articles about these things. One of the first articles I read in my newsfeed this morning has an explanation section that begins: “When photons are entangled, they behave as if they can instantly affect each other. This quantum mechanical phenomenon is essential to many quantum technology applications under development, such as quantum sensing, quantum communications or quantum computing.”

        In Scott Aaronson’s text on QC, as far as I can tell from skimming, MWI is mentioned only in the section that discusses QM interpretations. I can’t recall seeing it mentioned in any QC or cosmology article I’ve read. (One that was about QC or cosmology rather than about MWI in the first place.)

        So I searched for [MWI cosmology]. Google, knowing my research habits, gave me as the first two links the Wiki and SEP MWI pages. Both those pages touch on it, the Wiki page more so. It was the fourth link and the seventh link, both to PDF copies of papers. The latter was written by what almost seems an alternate copy of myself that did end up as a physicist. These papers are meaty enough I’ll give them their own comment below. (Eventually I’ll write a post about the alternate-self paper. Blew my mind!)

        “Certainly the idea of not just particles, but the whole universe in superposition is disconcerting.”

        It’s a mistake to think I’m led by my gut. I’m not bothered by being disconcerted; I’m bothered by non-physical assertions. Coincident matter. Thinning energy. No physical theory we know accounts for those.

        (The Kent paper gets into the sociological aspects of this. The objections are not emotional — they’re logical. In some cases, mathematical.)

        “Doesn’t mean someone won’t discover an actual collapse tomorrow and falsify it.”

        You focus on collapse, but a deep understanding of superposition and the quantum/classic divide may show that focus is misguided. It may be particularly misguided wrt to the SE (for reasons I’ve explained). The question should be what is happening physically, and I see MWI ignoring this by placing the SE at the center of the interpretation.

        At this point there doesn’t seem anywhere to go unless you want to defend Wallace from my objections or comment on the two papers I’ll describe below.

  • Wyrd Smythe

    This paper: Are there really many worlds in the “Many-worlds interpretation” of Quantum Mechanics? was very interesting to me because it reflects a view of MWI that I took many years ago (circa late 1990s) when I favored the MWI. It was published in January 2020 by Daniel Parrochia, a French philosopher.

    Essentially the paper argues that Everett, in his 1955 thesis and 1957 paper, never intended “many worlds” to be physically real and distinct. Instead it was the version DeWitt popularized in the 1970s that asserts these worlds. The paper claims the original view is that there is one quantum world that consists of superpositions. (This is somewhat similar to the Deutsch view, AIUI.)

    It’s a 33-page paper, clearly written, and with little math. Some relevant quotes…

    Regarding Chapter 5 of the Everett paper:

    So, as everyone can see, Everett never said that the reality to which we have access through the quantum wave function could be decomposed into a multiplicity of worlds (which, because of this “decomposition”, would be classical). He only said that, “as soon as the observation is performed”, the composite state is split into a superposition of different composite system-observer states.

    Quoting the rough draft of a letter Everett wrote in response to Lévy-Leblond in 1977:

    I have not done further work in this area since the original paper in 1955 (not published in entirely until 1973) as the “Many-Worlds interpretation etc.” This, of course, is not my title as I was pleased to have the paper published in any form anyone chose to do it in! I, in effect, had washed my hands of the whole affair in 1955. Far be it for me to look a gift Boswellian writer in the mouth! But your observations are entirely accurate (as far as I have read)

    Parrochia goes on to say about this:

    Everett actually says three things in essence: 1) I closed the file in 1955 and all that no longer interests me (I washed my hands); 2) One [DeWitt] wanted to name this theory “many worlds interpretation” and, of course, this is not the title I gave it. But I was not going to be choosy, since one [DeWitt] was republishing my work and, moreover, taking as much interest in myself as Boswell could have done by chronicling the life of Samuel Johnson; 3) That said, you are absolutely right.

    I think he might be over-emphasizing point #3 a teeny bit. To me Everett seems more to qualify his agreement parenthetically. The copy he actually sent to Lévy-Leblond reads, “In this case, your observations seem entirely accurate (as far as I read).” Everett also mentions his disengagement with the topic, so he may not have read the work carefully.

    Parrochia notes:

    The fact that Everett’s supposed multiverse (with giant superposition embracing branching worlds, etc.) is a key feature of his theory is all the less certain that Everett himself never said that.

    Parrochia continues with a discussion of tests of MWI and probability issues. I found it a clear and worthwhile discussion. How MWI connects with string theory was especially interesting.

    Near the end he says:

    In short, Everett’s thesis, as reviewed and corrected by DeWitt, lends to science-fiction. But we must not forget that there is science under fiction and that much of many-worlds theory is a “way of speaking”.

    I had to grin when he continues on to mention Dr. Sabine Hossenfelder:

    Going further risks exposing physicists to the harsh criticisms of Sabine Hossenfelder who argues that the physical community has wandered and that, for a long time, physics has taken a wrong turn with this idea of “multiverse”:

    He includes a quote from her book, Lost in Math.

    Next he explores quantum logic and points out what he sees as a fundamental mistake the DeWitt version of the MWI makes. It’s essentially that:

    ((|Ψ⟩_1 ∨ |Ψ⟩_2) ∧ R) = ((|Ψ⟩_1 ∧ R) ∨ (|Ψ⟩_2 ∧ R))

    (Where R is the interaction.) It’s false in quantum logic. (It’s true in classical logic.)

    He sums up:

    So we have a real problem to translate into natural language and make understandable to a human mind what the formulas of quantum mechanics say. This has philosophical consequences.

    The Kent paper (see below) says something similar.

    Parrochia concludes:

    But human mind continues to decompose all that in order to understand, and it is very difficult to find a true language between mathematics and fairy tales.

    So Parrochia supports what he sees as Everett’s true view — a single quantum reality of superposition, not what he sees as the DeWitt view of actual physically distinct worlds. As I mentioned at the top, this was my view of MWI back in the 1990s.

    One of the more salient points he makes (and Kent does also) is how badly language fails at describing any of this. Having only a language-based understanding is close to no understanding at all. One really does have to get into the math to really see anything clearly.

  • Wyrd Smythe

    This 1990 paper: Against Many-Worlds Interpretations blew me away. Sometime in the near future I want to do a full post about it. For now I’ll just say it’s like meeting the alternate version of me that went on to become a physicist (which was the original youthful plan).

    The author, Adrian Kent, is “a British theoretical physicist, Professor of Quantum Physics at the University of Cambridge, member of the Centre for Quantum Information and Foundations, and Distinguished Visiting Research Chair at the Perimeter Institute for Theoretical Physics. His research areas are the foundations of quantum theory, quantum information science and quantum cryptography. He is known as the inventor of relativistic quantum cryptography.” (Actually, probably way more impressive than I likely ever would have been.)

    I’m not typically big on provenance of ideas, but that’s an impressive credential. The acknowledgements at the end of the paper likewise impressed me:

    It is a pleasure to thank S. de Alwis, J. Anandan, W. Boucher, S. Carlip, P. Goddard, J. Halliwell, R. Penrose, R. Wald, F. Wilczek for interesting discussions and explanations of their views on various many-worlds interpretations. Thanks also to a referee for constructive criticisms, and for permission to quote from these. The hospitality of the Institute for Advanced Study is gratefully acknowledged. This work was supported by DOE grant DE-AC02-76ER02220.

    The paper is listed at at CERN and Inspire. Kent published it on arXiv in 1997 with an added forward.

    Which, in part, reads:

    There is clearly enormous appeal in the idea that, somehow, there simply cannot be any real problem in extending quantum theory from the Copenhagen interpretation to a fully satisfactory interpretation of quantum cosmology. It seems to me that faith in that idea now generally survives despite, rather than because of, the many different attempts by distinguished physicists to explain precisely how the quantum theory of the universe should be interpreted.

    Kent also says:

    Most people attracted by many-worlds ideas, however, tend to be working in other areas of physics. My impression is that most of these enthusiasts tend not to follow very closely the debates as to what, precisely, any given interpretation means and what its scientific implications are. And understandably so. The basic idea of a many-worlds interpretation seems superficially attractive: it is always unsettling to think that there may be an underlying problem which, at some stage, will have to be confronted, and it is reassuring to be told that there is an apparently unthreatening solution. Advocates of many-worlds interpretations tend, naturally, to stress the interest in their ideas, rather than the problems — which,of course, may not be such serious problems from their perspective. And it is hard to pick one’s way through the literature on interpretations of quantum theory, and easy to come away with the (unfortunately sometimes correct) impression that nothing very constructive comes of these debates.


    In the Introduction Kent takes up another point:

    The useful criticisms of MWI do not stem from an inability to accept the picture of multiple branching universes, nor do serious critics merely say that they do not see how to reduce macroscopic physics to an MWI. They assert either that particular physical facts are demonstrably not logically deducible from the axioms of MWI, or else they criticize the axioms proposed (usually on grounds of complexity or arbitrariness). Though it does not seem to be widely recognized, these criticisms have had a substantial effect: modern opponents of state vector reduction, recognizing that Everett’s original argument is erroneous, have adopted increasingly sophisticated formulations.

    A point I’ve had to make many times. My objections to the MWI have nothing to do with emotional or gut responses. It’s when I sit down and think through all the implications that paradoxes arise.

    Another important point:

    In seeking this sort of clarification, we are not asking for anything extraordinary. Simply, MWI proponents need to specify precisely the mathematics involved in setting up their theory, to state explicitly which of the quantities they define are supposed to be physical, and to explain (in principle, not necessarily in detail) how our observations are to be described in terms of these physical quantities. Examples of theories which satisfy these criteria are the standard presentations of Maxwell’s electrodynamics and general relativity. Here the electric and magnetic fields, matter and charge densities, space-time manifold and metric are taken to be physical quantities, while the gauge potentials and the various coordinate systems on the manifold are only part of the mathematical formalism.

    That’s a very good way to put it. As Kent says just before this:

    Secondly, we have tried to clarify the logical structure of the MWI, and so to pinpoint their essential problems. This involves the extraction and formalization of what seem to us the fundamental assumptions in the original papers. The attempt may not have entirely succeeded. But we are convinced that the procedure is justified, and in fact that axiomatization should have been insisted upon from the beginning. For any MWI worth the attention of physicists must surely be a physical theory reducible to a few definite laws, not a philosophical position irreducibly described by several pages of prose. Moreover, once an axiomatization is achieved, the irrelevance of much of the rhetorical argument surrounding MWI is exposed, and any given MWI’s (de)merits become much clearer

    Again, could not have said it better myself. I especially agree with, “For any MWI worth the attention of physicists must surely be a physical theory reducible to a few definite laws, not a philosophical position irreducibly described by several pages of prose.”

    This is, perhaps, my biggest objection to MWI proponents: It’s just words, and words aren’t enough in quantum mechanics.

    Kent examines a variety of MWI versions in detail, especially Everett’s, Graham’s, and DeWitt’s. Kent also touches on Cooper-Van Vechten, Hartle, and Farhi-Goldstone-Gutmann. In all cases, Kent seeks to derive axioms from their expressed views. He then examines the consequences of the axioms, and identifies issues with them.

    Kent also comments on Geroch and Deutsch.

    In the discussion section, Kent writes:

    Everett’s original attempt at a many-worlds interpretation was an interesting and well-motivated idea. Everett wanted simultaneously to simplify the axioms of quantum mechanics and to produce a realist physical theory (one which gives an explicit mathematical description of an external reality). However, it is clear that he did not succeed. Successive advocates of many-worlds interpretations, while usually claiming merely to clarify Everett’s position, have substantially abandoned his goals. No proponent of MWI (as far as we are aware) has actually produced a complete set of axioms to define their physical theory. It seems that any such axioms must involve the extremely complicated notion of a measuring device, or that the physical reality they define will be one without any notion of timelike trajectory (and so, as Bell [8,9] has pointed out, not obviously consistent with special relativity), or both.

    He continues:

    All of this suggests that Everett’s original program has degenerated, apparently beyond repair. Although the logical possibility of an MWI consistent with known physics cannot completely be excluded, it seems that the defining axioms of such a theory would have to be extremely ugly and arbitrary. How then can the continued attraction of MWI be accounted for? The question is really one for future historians of science, but perhaps we can suggest two possible answers.

    He answers that:

    Firstly, the very failure of MWI proponents to axiomatize their proposals seems to have left the actual complexity of realistic MWI widely unappreciated. It may thus possibly be tempting for MWI advocates to assume that there is no real problem; that Everett’s detractors either have not understood the motivation for, or merely have rather weak aesthetic objections to, his program.


    Secondly, MWI seem to offer the attractive prospect of using quantum theory to make cosmological predictions. The trouble here is that if an MWI is ultimately incoherent and ill-founded, it is not clear why one should pay attention to any quantum cosmological calculations based on it. Perhaps the right attitude should be to explore with interest, but to take any results that depend on MWI as no more than suggestions — not to be taken seriously without independent proof.

    And I can’t add a thing to that!

    His conclusions even align with mine:

    Our guess is that the postulates of standard quantum mechanics (I-V above) are not an ultimate description of nature; the ultimate description is given by postulates as yet unknown, which are probably framed in a different mathematical language. However, insofar as the language of state vectors in a Hilbert space applies to physics, postulates I-V are a good approximation to the behaviour of microscopic systems. In particular, the state vector reduction postulate is an approximate description of what happens to a microscopic system when it interacts with a macroscopic one. For macroscopic systems, either the postulates do not hold (the Schrödinger equation ceases to be a good approximation and the postulate of state vector reduction ceases to be well-defined) or the whole language of state vectors is inapplicable. The distinction between microscopic and macroscopic should be one of degree not of kind, and should follow from the unknown fundamental postulates.

    Exactly what I’ve been saying for a long time now.

    So, all in all, an amazing paper from my perspective. It supports everything I’ve been saying here. It even touches on two-slit experiments. It also gets into what I’ve been saying about wave-function indefiniteness — it’s a bigger problem that MWI proponents seem to acknowledge.

  • SelfAwarePatterns

    Some examples of cosmologists who are (or were) Everettians: Sean Carroll, Mag Tegmark, Stephen Hawking, Alan Guth, Lawrence Kraus, Martin Rees, and Leonard Susskind. (Some of these guys are agnostic about the other worlds. They’re real attachment is to the collapse free wave mechanics.)

    Matt O’Dowd gives a good overview of why Von Neumann / Wigner variants of collapse interpretations are problematic for cosmology:
    If you think about it, the black hole information paradox is meaningless under a collapse interpretation, where most of the information in the universe is constantly being lost.

    David Deutsch is the most prominent quantum computing theorist I can think of who supports Everett. Aaronson has talked favorably about the MWI in interviews, although he seems to stay very aware it’s only viable, not yet reliable info. (A stance I fully agree with.) (It doesn’t surprise me he wouldn’t impose his opinion on a class. A mark of a quality teacher.)

    That said, I do think any interpretation with wave function realism is compatible with quantum computing, so that would include objective collapse theories like GRW.

    We’ve talked about the “many worlds” and “multiverse” terminology before, and how it wasn’t Everett’s, at least not in his original papers. The thing to remember is that the formalism of Everettian physics is the raw quantum formalism. For most versions, the rest amounts to different ways of talking about it.

    Some of those ways, like DeWitt’s and Deutsch’s, are more sensationalist than others, and there are physicists who are fine with core Everettianism but dislike multiverse talk. (Woit in his latest post and thread, seems sympathetic with that stance.) I can see their point, because it gives people the wrong idea. All the brief descriptions I initially read of it implied whole new universes being instantly created, which I found ridiculous. I had to unlearn that view to find the theory at all plausible.

    On the other hand, I don’t know that we’d even be talking about this theory if DeWitt hadn’t promoted it the way he did. Everett obviously understood this.

    But I do sometimes wonder if it might be taken more seriously if it was called by Everett’s name: “Universal Wavefunction”, or Wheeler’s: “Relative State Formulation”. But most Everettians seem to have settled on variants of “Everettian Quantum Mechanics” or the “Everett Interpretation”.

    On the other paper, I wonder if the authors ever tried to make sense of Niels Bohr’s writing or any of the other Copenhagen rationalizations. The paper seems extremely partisan.

    • Wyrd Smythe

      Extremely partisan! 😀 I wondered how you would disdain such a thorough and detailed analysis.

      At this point I think we’ve said as much as we can. We seem to have established we disagree on just about every aspect of this.

  • The nature of quantum nonlocality – SelfAwarePatterns

    […] physics has been on my mind again lately, somewhat triggered by a recent conversation with Wyrd Smythe on his blog. I’ve always known quantum nonlocality has nuances, but stuff I […]

And what do you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: