The Big Picture

Home > Other > The Big Picture > Page 63
The Big Picture Page 63

by Carroll, Sean M.


  02

  we think about quantum measurement. Penrose has some specific ideas

  03

  about what those alterations might be— quantum gravity is involved, and

  04

  filamentary structures in the brain called microtubules— but the upshot is

  05

  that the wave functions of structures in our brains collapse in just the right

  06

  way to grant human beings powers of insight and cognition that computers

  07

  will never achieve.

  08

  There are a number of objections one could raise, and people have had

  09

  fun raising them against Penrose for years now. The best ones center on the

  10

  leap from “Human cognition doesn’t work like a formal mathematical sys-

  11

  tem” to “The human brain doesn’t obey the known laws of physics.” What

  12

  we call “thinking” is a way of talking about a very high- level emergent phe-

  13

  nomenon. It may emerge out of underlying processes that are absolutely

  14

  rigid and logical, and yet itself not show those characteristics very much at

  15

  all. Indeed, rigid logic (or even the ability to multiply big numbers accu-

  16

  rately) is something that human beings are notoriously bad at. Our thoughts

  17

  leap around, we make mistakes, we have hunches. The fact that we can

  18

  reach conclusions that wouldn’t be reached by a specific formal system

  19

  doesn’t seem particularly surprising.

  20

  Gödel’s Incompleteness Theorem doesn’t quite say there are true state-

  21

  ments that can’t be proven. Rather, it says that such statements exist for any

  22

  consistent formal system. How do we know that some particular set of axi-

  23

  oms defines a consistent system? Or— putting it another way— how can we

  24

  be sure that we are accurately “perceiving” the truth of Gödel’s self-

  25

  referential sentences?

  26

  As Scott Aaronson has pointed out, it’s more accurate to say that we

  27

  believe certain systems are consistent, though Gödel has shown that we can

  28

  never prove it. If we allow a computer to assume that the system is consis-

  29

  tent, it would have no trouble at all proving statements like “This statement

  30

  cannot be proven.” (Proof: if it could be proven, the system would be incon-

  31

  sistent!) He quotes Alan Turing: “If we want a machine to be intelligent, it

  32

  can’t also be infallible. There are theorems that say almost exactly that.”

  33

  Humans certainly satisfy the criterion of not being infallible.

  34

  Putting on our Bayesian hats, the fact that the convoluted minds of hu-

  35S

  man beings naïvely seem to be able to perceive truths that can’t be directly

  36N

  proven by completely rigorous computer programs doesn’t seem nearly

  370

  Big Picture - UK final proofs.indd 370

  20/07/2016 10:02:55

  A R E PhO t O n S C O n S C I Ou S ?

  strong enough to warrant modifying our best understanding of quantum

  01

  mechanics. Especially because the use to which such modifications are be-

  02

  ing put has nothing directly to do with the mysteries of quantum mechan-

  03

  ics themselves— it’s just a way to grant powers of insight and cognitive

  04

  wizardry to the human brain. And at the end of the day, there’s nothing

  05

  about the brain’s ability to see the truth of unprovable statements that helps

  06

  us understand the Hard Problem, the issue of inner mental experiences. If

  07

  you think the Hard Problem is hard, quantum mechanics is unlikely to

  08

  help you; if you think it’s not so bad, you probably don’t feel the need to

  09

  change the laws of physics to help us understand the brain.

  10

  11

  12

  13

  14

  15

  16

  17

  18

  19

  20

  21

  22

  23

  24

  25

  26

  27

  28

  29

  30

  31

  32

  33

  34

  S35

  N36

  37 1

  Big Picture - UK final proofs.indd 371

  20/07/2016 10:02:55

  01

  02

  03

  43

  04

  05

  What Acts on What?

  06

  07

  08

  09

  10

  11

  12

  13

  14

  15

  16

  17

  18

  the idea that we are part of the natural world can lead to a sense of

  profound loss if the reasons and causes for our actions aren’t what

  we thought they were. We’re not human beings, equipped with in-

  tentions and goals, so the worry goes; we’re bags of particles mindlessly

  19

  bumping into one another as time chugs forward. It’s not love that will keep

  20

  us together, it’s just the laws of physics. A version of this concern was ar-

  21

  ticulated by philosopher Jerry Fodor:

  22

  23

  If it isn’t literally true that my wanting is causally responsible

  24

  for my reaching, and my itching is causally responsible for my

  25

  scratching, and my believing is causally responsible for my say-

  26

  ing . . . if none of that is literally true, then practically every-

  27

  thing I believe about anything is false and it’s the end of the

  28

  world.

  29

  30

  Don’t worry! It’s not the end of the world.

  31

  We live in a reality that can be fruitfully talked about in many differ-

  32

  ent ways. We have an extravagant assortment of theories, models, vocabu-

  33

  laries, stories, whatever you prefer to call them. When we speak about

  34

  a human being, we can describe them as a person with desires and tenden-

  35S

  cies and inner mental states; or we can describe them as a collection of

  36N

  372

  Big Picture - UK final proofs.indd 372

  20/07/2016 10:02:55

  W h At AC t S O n W h A t ?

  biological cells interacting via electrochemical signals; or we can describe

  01

  them as an agglomeration of elementary particles following the rules

  02

  of the Core Theory. The question is, how do we fit these different stories

  03

  together? In particular, what acts on what? Does the existence of

 
04

  the particle- physics description, in which “causality” is nowhere to be

  05

  found, imply that it is illegitimate to talk about scratching being caused by

  06

  itching?

  07

  The poetic- naturalist answer is that any of the stories we have stands or

  08

  falls on its own terms as a description of reality. To evaluate a model of the

  09

  world, the questions we need to ask include “Is it internally consistent?,” “Is

  10

  it well- defined?,” and “Does it fit the data?” When we have multiple distinct

  11

  theories that overlap in some regime, they had better be compatible with

  12

  one another; otherwise they couldn’t both fit the data at the same time. The

  13

  theories may involve utterly different kinds of concepts; one may have par-

  14

  ticles and forces obeying differential equations, and another may have hu-

  15

  man agents making choices. That’s fine, as long as the predictions of the

  16

  theories line up in their overlapping domains of applicability. The success

  17

  of one theory doesn’t mean that another one is wrong; that only happens

  18

  when a theory turns out to be internally incoherent, or when it does a bad

  19

  job at describing the observed phenomena.

  20

  Developing a theory of human thought and behavior in terms of neu-

  21

  ral signals or interacting particles doesn’t in any way imply that your

  22

  wanting is not responsible for your reaching. There is no obstacle to

  23

  that kind of vocabulary of desire and intentionality being “true,” as long

  24

  as its predictions are compatible with those of other successful vocabularies.

  25

  It’s possible that what Fodor means by “literally true” is something like

  26

  “an essential element of every possible description of nature,” or perhaps “of

  27

  our best and most comprehensive description of nature.” In other words,

  28

  there can’t exist any successful vocabulary that doesn’t include “wanting”

  29

  and “believing” as fundamental concepts. In that case, it is not literally

  30

  true— the physical and biological descriptions of human beings are per-

  31

  fectly adequate on their own terms, and don’t invoke concepts like wants

  32

  and beliefs.

  33

  But that’s an unnecessarily constraining notion of “literally true.” Ther-

  34

  modynamics and the fluid description of air didn’t stop being true once we

  S35

  N36

  373

  Big Picture - UK final proofs.indd 373

  20/07/2016 10:02:55

  T H E B IG PIC T U R E

  01

  discovered atoms and molecules. Both ways of talking are true. Likewise,

  02

  human thoughts and intentions haven’t disappeared just because we obey

  03

  the laws of physics.

  04

  •

  05

  06

  This issue seems more complicated than it is because of an understandable

  07

  tendency, in a world described by multiple distinct but mutually compatible

  08

  stories, to jumble up the concepts of one story with those of another— to

  09

  cross the lines separating distinct ways of talking.

  10

  Rather than acknowledging that there is one way of talking about the

  11

  world in terms of the quantum fields and interactions of the Core Theory,

  12

  and another way in terms of electrochemical signals traveling between cells,

  13

  and yet another way in terms of human agents with desires and mental

  14

  states, we fall into the trap of using multiple vocabularies at the same time.

  15

  When told that every mental state corresponds to various physical states of

  16

  one’s brain, one wants to complain, “Do you really think the reason why I’m

  17

  scratching is only because of some synaptic signaling, and not because I feel

  18

  an itch?” The complaint is misplaced. You can describe what’s happening in

  19

  terms of electrochemical signals in your central nervous system, or in terms

  20

  of your mental states and the actions they cause you to perform; just don’t

  21

  trip up by starting a sentence in one language and attempting to finish it in

  22

  another one.

  23

  One of the most common arguments against Cartesian dualism (or

  24

  mental properties that influence physical ones) is causal closure of the physi-

  25

  cal. The laws of physics as we know them— the Core Theory, in the domain

  26

  we’re interested in— are complete and self- consistent. You give me a quan-

  27

  tum state of a system, and there are unambiguous equations that will tell

  28

  me what it will do next. (We’ve written down one such equation in the

  29

  Appendix.) There is no ambiguity, no secret fudge factors, no opportunity

  30

  for differing interpretations of what is happening. If you give me the precise

  31

  and complete quantum state corresponding to “a person feeling an itch,”

  32

  and I have the calculational abilities of Laplace’s Demon, I could predict

  33

  with extraordinary accuracy that the quantum state will evolve into a dif-

  34

  ferent state corresponding to “a person scratching themselves.” No further

  35S

  information is needed, or allowed.

  36N

  374

  Big Picture - UK final proofs.indd 374

  20/07/2016 10:02:55

  W h At AC t S O n W h A t ?

  •

  01

  02

  In chapter 13 we discussed the idea of “strong emergence,” according to

  03

  which the behavior of a system with many parts is not reducible to the ag-

  04

  gregate behavior of all those parts. A related idea is downward causation:

  05

  behavior of the parts is actually caused by the state of the whole, in a way

  06

  not interpretable as due to the parts themselves.

  07

  Poetic naturalists tend to view downward causation as a deeply mis-

  08

  guided idea. Then again, they view upward causation as equally misguided.

  09

  “Causation,” which after all is itself a derived notion rather than a funda-

  10

  mental one, is best thought of as acting within individual theories that rely

  11

  on the concept. Thinking of behavior in one theory as causing behavior in

  12

  a completely different theory is the first step toward a morass of confusion

  13

  from which it is difficult to extract ourselves
.

  14

  It’s certainly possible that behavior in coarse- grained macroscopic theo-

  15

  ries might be entailed by features of more comprehensive theories, and we

  16

  certainly want them to be consistent with such theories when the descrip-

  17

  tions overlap. We might even, as long as we’re careful, say that features of an

  18

  underlying theory can help explain features of an emergent one. But we get

  19

  in trouble if we try to say that phenomena in one theory are caused by phe-

  20

  nomena in a different one. I know that I cannot use my mental powers to

  21

  reach across space and bend spoons, since the fields and interactions of the

  22

  Core Theory don’t accommodate that kind of capacity. But I can describe

  23

  that feature purely in the macroscopic language: human beings don’t pos-

  24

  sess the power of telekinesis. The microscopic explanation might aid my

  25

  understanding, but it’s not a necessary part of how I talk about human- scale

  26

  behavior.

  27

  And the converse, downward causation of human- scale properties influ-

  28

  encing the microscopic behavior of particles, is misguided. A standard ex-

  29

  ample is the formation of snowflakes. Snowflakes are made of water

  30

  molecules, interacting with other molecules to form a crystalline structure.

  31

  There are many possible structures, determined by the initial configuration

  32

  of the seed from which the snowflake grows. Therefore, it is claimed, the

  33

  macroscopic shape of the snowflake is acting “downward” to determine the

  34

  precise location of individual water molecules.

  S35

  N36

  375

  Big Picture - UK final proofs.indd 375

  20/07/2016 10:02:55

  T H E B IG PIC T U R E

  01

  It’s bad form to mix vocabularies in such a vulgar way. Water molecules

  02

  interact with other water molecules, and other molecules in the air, in pre-

  03

  cise ways that are specified by the rules of atomic physics. Those rules are

  04

  unambiguous: you tell me what other molecules any individual water mol-

  05

  ecule is interacting with, and the rules will say precisely what will happen

  06

  next. The relevant molecules may be part of a larger crystalline structure,

  07

  but that knowledge is of zero import when studying the behavior of the

  08

  water molecule under consideration. The environment in which the mole-

  09

  cule is embedded is relevant, but there is no obstacle to describing that en-

  10

 

‹ Prev