of 39 /39
Relativistic quantum theory: particles or fields? Deeper troubles: non-locality and the MP Relativistic Quantum Physics Christian Wüthrich Université de Genève 4th International Summer School in Philosophy of Physics Saig, 21 July 2016 Christian Wüthrich Relativistic Quantum Physics

Relativistic Quantum Physics

Embed Size (px)

Text of Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Relativistic Quantum Physics

    Christian Wthrich

    Universit de Genve

    4th International Summer School in Philosophy of PhysicsSaig, 21 July 2016

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    My sources (without detailed credit)

    Anthony Duncan (2012). The Conceptual Framework of Quantum Field Theory.OUP: Oxford.

    Lawrence P Horwitz (2015). Relativistic Quantum Mechanics. Springer:Dordrecht.

    Tom Lancaster and Stephen J Blundell (2014). Quantum Field Theory for theGifted Amateur. OUP: Oxford.

    Lewis H Ryder (1985). Quantum Field Theory. CUP: Cambridge.

    Steven Weinberg (1995). The Quantum Theory of Fields. Volume 1: Foundations.CUP: Cambridge.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    John S Bell on what the real problem is

    For me... this is the realproblem with quantum theory:the apparently essential con-flict between any sharp formu-lation and fundamental relativ-ity. That is to say, we have anapparent incompatibility, at thedeepest level, between the twofundamental pillars of contem-porary theory.

    J S Bell. Speakable and unspeakable in quantummechanics. In his eponymous paper collection, CambridgeUP (2004): 172.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    W B Yates

    The innocent and the beautifulHave no enemy but time

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Inspired by W B Yates

    The particulate and the quantum mechanicalHave no enemy but spacetime

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    The superficial problem and its resolution

    W G Unruh. Minkowski space-time and quantum mechanics. In V. Petkov (ed.), Minkowski Spacetime: AHundred Years Later, Springer (2010): 133-148.

    In the setting of a quantum-mechanical theory of particles, theonly Lorentz-invariant interactions which conserve energy andmomentum are contact interactions between point particles(Unruh 2010, 137).

    More generally, the tension between such a theory and SR canbe seen when comparing the wave function (t , ~x) (where t is auniversal parameter, ~x are eigenvalues for position operator) andscalar field (t , ~x) (where t , ~x are coordinates of spt point).

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Wave function vs. scalar fields

    W G Unruh, Minkowski space-time and quantum mechanics, in V. Petkov (ed.), Minkowski Spacetime: AHundred Years Later, Springer (2010): 133-148.

    Compare: wave function for two particles (t , ~x , ~y) vs. pointinteractions of two (identical) scalar fields (t , ~x) (t , ~x)

    The single time t needed for the wave function of two particlesentail a need for a privileged foliation of spacetime; not so forfields.

    Not an argument, just an indication. Nevertheless:

    early recognition that a quantum field theory is needed to easethe tension with SR

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    The beginnings of relativistic quantum physics

    1925: Schrdinger considers, but rejects, relativistic waveequation

    1925: first quantum field theory (QFT) appears in the part writtenby Jordan in the Drei-Mnner-Arbeit with Born and Heisenberg

    1926: Klein and Gordon (separately) propose their eponymousequation as description of relativistic electrons

    1928: Dirac proposes his eponymous equation as description ofrelativistic electrons

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Overheard at the 1927 Solvay conference:

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Klein-Gordon equation for a single particle

    E = p2/2m E2

    c2 p p = m2c2

    E i~ t ,p i~.

    ~22m

    2 = i~t (m2) = 0

    (in natural units)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    The changing fate of the probability interpretation

    ~22m

    2 = i~t (m2) = 0

    (in natural units)

    = probability density = i~2m

    ( t

    t

    )

    positive definite not positive definite

    KG equation is second order, and /t can be fixedarbitrarily at a time, and so may become negative

    interpretation of as probability density must be given up

    Also, vanishes for a real ; correct interpretation of : chargedensity (complex corresponds to charged particle)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    The Feynman-Stckelberger interpretation

    Another problem with the KG equation is that the relativisticdispersion relation (regarded as equation for E), admits negativeenergy solutions:

    E = (m2c4 + p2c2)1/2.

    Feynman-Stckelberger: negative energy states as positiveenergy antiparticles, i.e., particles moving backward in time (withmomenta in opposite direction)

    general solution of the KG equation is a superposition of twostates: incoming particle and outgoing antiparticle

    general solution cannot be interpreted as description of singleparticle

    better interpretation of KG equation: describes dynamics offields whose excitations are spinless particles

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Further considerations against a single-particleinterpretation

    TD Newton and E P Wigner. Localized states for elementary systems. Reviews of Modern Physics 21(1949):400-406.

    Newton and Wigner add another reason that a relativisticsingle-particle QM is impossible:

    for a zero-spin particle described by KG equation, there is asmall but non-vanishing probability of finding the particle whichwas originally located at the origin outside its forward lightcone.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Further considerations against a single-particleinterpretation

    Furthermore,

    particle with mass m squeezed into a box smaller than itsCompton wavelength has a position uncertainty of x ,and so p h/ = m

    Given this large energy (greater than its mass), SR suggeststhat particle/antiparticle pairs are being produced.

    Thus, on average, the box contains more than one particle,challenging the single-particle interpretation.

    All these considerations force abandoning a straightforwardparticle concept in relativistic quantum physics.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Guy Fawkes

    A desperate disease requires a dangerous remedy.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Fields!

    According to Lancaster and Blundell, QFT at core is the idea that[e]very particle and every wave in the Universe is simply an excitationof a quantum field that is defined over all space and time. (1)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    What is a field?

    A field is some kind of machinethat takes a position in space-time, given by the coordinate x,and outputs an object represent-ing the amplitude at that point inspacetime. Here the output is thescalar (x) but it could be, for ex-ample, a vector, a complex num-ber, a spinor or a tensor. [Quan-tum fields are captured by oper-ators which are locally defined,e.g., (x).]

    (Caption in Lancaster and Blundell, 2)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Local relativistic quantum field theory

    Local relativistic QFT is based on three principles (Duncan, 57f):

    1 Quantum mechanics: linear superposition of amplitudes,probability interpretation of these amplitudes (squared), andunitary evolution of quantum state, implementing the dynamicsof the theory

    2 Special relativity: symmetry of Lorentz invariance

    3 Clustering: insensitivity of local processes to the distantenvironment (58); in context of relativistic theories, thisbecomes locality or microcausality, which ensures, e.g.,absence of superluminal signalling

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Consequences of the basic principles

    forces resorting to approximative schemes (bad)

    enforces constraints on Hamiltonian (good)

    Two more important consequences:1 gives explanation of existence of antimatter, with each

    particle possessing an antiparticle of equal mass andopposite additive quantum numbers

    2 spin-statistics theorem, clarifies contrasting symmetryproperties of the wavefunctions of particles of integer(bosonic) and half-integer (fermionic) spin

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Antiparticles(From Duncan (2012,59f), from Weinberg 1972, going back to Feynman 1949)

    Consider a process where a proton (P) emits a positive pion (+)at spacetime event x , which then travels to y to be absorbed bya neutron (N):

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    locality here amounts to interaction between neutron andproton via local emission and absorption events of thirdintermediary particle

    indeterminacy of position and velocity in QM permits for x and yto be space-like related

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    SR: Lorentz boost of left frame to frame on right changestemporal order of x and y so that y now precedes x

    observer in right frame: negative pion () emitted by neutron Nat y is later absorbed by proton P at x

    intimate association between spatiotemporal reflection andparticle-antiparticle interchange characteristic of local theoriesand exemplified in the [CPT] theorem (60)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Troubles for the (single) particle interpretationFields

    Locality and superluminality

    But dont the s travel superluminally and thus violate SR?

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    At the heart of the tension: locality in axiomatic QFT

    In axiomatic QFT, a local action principle is usually observed,

    where local action means something like no physical actioncan propagate faster than the speed of light, or at least thatthere is no transmission of statistically measurable properties atfaster than light speed, i.e., no signalling.

    local actions in one spacetime region cannot changemeasurable properties in regions spacelike to it

    operators in spacelike related regions commute

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    At the heart of the tension: locality in axiomatic QFT

    Locality in the Haag-Kastler axioms means that algebras living inspacelike separated regions commute, as follows:

    Axiom (Haag-Kastler Locality)

    Given an algebra A(O) defined over a spacetime region O M, withO M denoting the set of spacetime points spacelike separatedfrom every point in O and A the set of operators that commute withevery operator in A (the commutant of A), then

    A(O) A(O).

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Algebra of complement as subalgebra of commutant

    )'O(A )'O(A

    i M;h

    'O 'OO

    )O(A

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Proof by construction?

    Does relativistic QFT give us a proof by construction thatquantum physics and SR are compatible?

    Superficially, it certainly seems so:1 Klein-Gordon and Dirac equations are relativistically

    invariant2 quantum states transform under Lorentz transformations as

    they should3 conditions such as Haag-Kastler Locality demand that

    superluminal action is ruled out

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Tim Maudlin: claims like (3) are empty because...

    Tim Maudlin. Quantum Non-Locality and Relativity. Blackwell Publishing (22002, 32011).

    Spatial separation would be a per-fect insulator from causal influ-ences if no influence could go fasterthan light. Since Aspects experi-ments shows no such insulating ef-fects, some direct causal connec-tions must exist between space-likeseparated events. (64) ... [R]eliablereproduction of the quantum statis-tics demands superluminal causa-tion and superluminal informationtransformation. (194)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    A sufficient condition for causation: counterfactual support

    One immediately wants to ask: based on what analysis ofcausation does Maudlin make these pronouncements?

    As he insists (126), a full analysis is unnecessary for presentpurposes; according to him, it suffices to have the followingsufficient condition for there to be causation:

    ConditionGiven two local physical events A and B, if B would not haveoccurred had A not occurred (or if B would have been different had Abeen different) then A and B are causally implicated with each other.(128) (Although one cant absolutely determine one to be the causeand the other the effectthey are spacelike separated after all.)

    only first approximation: relational nomic connections excluded(to avoid counting Cambridge change as causal implication)

    questionable: really sufficient if no direction of causal influencecan be determinedeven in principle?

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Giving short shrift to Maudlin

    In his detailed analysis of the possibly causal connection in Bellcorrelations in Chapter 5, Maudlin seems to make much weakerclaims than either before or after that chapter: the connection isonly one of causal implication, not causation.

    (Causal implication may obtain between events neither of which is the cause ofthe other; cf. 129)

    dilemma: either causal implication significantly weaker thancausation or not

    If it is, then it arguably doesnt violate SR, but its unclearhow this could satisfy our appetite for a causal explanationof the correlations; so its harmless, but impotent.If it is not, then that thirst is quenched, but on pains ofviolating SR.

    either we disappoint our causal intuitions or we reject SRSo lets see just how much we can get...

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Analyzing factorizability

    Jeremy Butterfield. Annals of the New York Academy of Sciences 755 (1995): 768-785.

    Butterfield (1995): Landau, Summers and Werner haveestablished (between 1985 and 1988) that the Bell inequality isgenerically violated in AQFT.

    In fact, there is a perfectly respectable realist approach to Bellcorrelations in non-relativistic QM, which gets encoded inaxiomatic QFTalbeit not in terms of causally efficacious localbeables.

    To see this, consider the usual EPR-Bohm setup with a singletstate: |(12) | (1) | (2) | (1) | (2)

    Main premise to derive Bell inequality:

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Factorizability condition

    p(12) (xL, xR |L,R) = p(1) (xL|L) p

    (2) (xR |R),

    where L and R are the measurement settings and xL and xR theoutcomes for the two wings, respectively.

    Factorizability is equivalent to the conjunction of two conditions:

    Outcome independence (OI)

    p(1) (xL|L,R, xR) = p(1) (xL|L,R)

    p(2) (xR |L,R, xL) = p(2) (xR |L,R)

    Parameter independence (PI)

    p(1) (xL|L,R) = p(1) (xL|L)

    p(2) (xR |L,R) = p(2) (xR |R)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Axiomatic QFT: violating OI

    Mikls Rdei and Stephen Summers, Foundations of Physics 32 (2002): 335-355.

    As Butterfield (1995) shows, PI is encoded in Haag-KastlerLocality (or some similar axiom).

    violation of OI, i.e. of separability

    Important: OI and PI are statements of probabilisticindependence, their denial falls short of a commitment to acausal dependence.

    It seems as if Bells theorem in itself does not entail that there beSR-violating causation.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    The dynamical threat: the MP

    Maudlin insists that these features [Lorentz invariance,non-separability?] could secure the credentials only of a theorywhich has no wave collapse. (194)

    I take him to be saying that even if you accepted that thenon-locality is taken care of in QFT, dynamically the relativisticcompliance of QFT is not assured.

    In other words, a relativistically accepted solution of the MP isrequired!

    And the MP certainly rears its face in QFT...

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    Taking measureConsider a measurement setup in the rest frame of a lab:

    post-measurement

    pre-measurement

    ??

    +

    SR OR x

    0t

    1t

    t

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    The MP in QFTJ Barrett. Wigners friend and Bells local field beables. Electronic J of Theoretical Physics. Forthcoming.

    field F : both state of system S in spatial region RS and state ofobserver O in spatial region RO|S0 (t): state of F is zero in RS at t|S+(t): state of F is nonzero in RS at tmutatis mutandis for region RO (subscript r for ready state)

    interaction Hamiltonian between local fields such that

    |Or (t0) |S0 (t0) |O0 (t1) |S0 (t1)|Or (t0) |S+(t0) |O+(t1) |S+(t1)

    Suppose field state in RS at t0 is |S(t0) = |S0 (t0)+ |S+(t0).unitary dynamics field state in RS and RO at t1 will be

    |O0 (t1) |S0 (t1)+ |O+(t1) |S+(t1)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    But in standard understanding, field state in RO must beseparable from state in RS after the interaction for there to be adeterminate record of the observers measurement, i.e. it musteither be |O0 (t1) (with probability ||2) or |O+(t1) (withprobability ||2).

    The moral is that insofar as there aredeterminate physical measurementrecords, if the quantum-mechanicalstate is taken to be complete, thenfield theory gets the dynamics of mea-surement wrong; and insofar as thereare determinate physical measure-ment records and the unitary dynam-ics is right, the quantum-mechanicalstate cannot be complete. (3)

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    A solution to the MP with relativistic credentials?

    So, the MP must be addressed in QFT as well. And this is thedynamical threat to the compatibility between quantum physicsand relativity: to solve the MP without violating SR.

    A solution to the MP must also offer an account of the Bellcorrelations, e.g.

    Is OI or PI violated, and how?What is the causal goings-on in a EPR-Bohm situation?Are there causal arrows at all, and if so, are they directed?

    This can only be denied on pain of rejecting realism.

    MP in non-relativistic QM la Maudlin (1995): either collapsetheory or hidden-variables theory or many-worlds theory

    Tim Maudlin. Three measurement problems. Topoi, 14 (1995): 7-15.

    Christian Wthrich Relativistic Quantum Physics

  • Relativistic quantum theory: particles or fields?Deeper troubles: non-locality and the MP

    Locality and non-locality in QFTThe real threat: the measurement problem

    MP in QFT/relativistic QM1 collapse:

    only proposal is rGRWf (Tumulka 2006)is inchoate, so far no standard modelor even interactions (aproblem for AQFT too)

    2 hidden variables:some field theoretic attempts (Bohm & Hiley; alt: Goldstein et al.)source of difficulty: Bohmian mechanics picks poison of causalmechanism over compatibility with SRBohmian mechanics violates PI and maintains OI, (axiomatic)QFT the other way around

    3 many worlds:most promising: seems to only involve elements which can bemade Lorentz invariantproblem: not that its ontologically profligate, but to show howsemblance of collapse or indeterministic statistics arises (how dowe get probabilities, Born rule)

    R Tumulka. A relativistic version of the Ghirardi-Rimini-Weber Model. Journal of Statistical Physics 125(2006): 821-840.

    Christian Wthrich Relativistic Quantum Physics

    Relativistic quantum theory: particles or fields?Troubles for the (single) particle interpretationFields

    Deeper troubles: non-locality and the MPLocality and non-locality in QFTThe real threat: the measurement problem