5D Library › Interoceptive Decoupling, Salience-Network Switching and Autonomic Visceral Restoration · 中文

Chapter 3. The Non-Judgmental Imagery Stream: Trading Errors, Not Gaining Information

What operation is “not interpreting”

The third link of the original claim is the most concrete: see the inner picture without asking what it means. There are corresponding concepts in the meditation literature. Bishop and colleagues’ operational definition of mindfulness has two components: self-regulation of attention, and an orientation of curiosity, openness and acceptance toward present experience (Bishop et al. 2004). Hölzel and colleagues decompose the mechanisms into attention regulation, body awareness, emotion regulation and change in the view of self (Hölzel et al. 2011). “Decentering” — treating a thought as a thought rather than as a fact — is a core concept in cognitive therapy.

But to test it, it has to be written as an operation. This chapter writes “interpretation” as a gain: an inner event supplies a piece of evidence, and semantic processing amplifies that evidence and drives it toward a decision.

What a gain changes

Let an inner event yield evidence e∼𝒩(μ,1)e \sim \mathcal{N}(\mu, 1) when the situation is in fact dangerous and e∼𝒩(0,1)e \sim \mathcal{N}(0, 1) when it is not. Let semantic processing be a gain gg and the decision criterion cc. Then

hit rate=1−Φ(c−gμg),false alarms=1−Φ(cg),d′=μ.\text{hit rate} = 1 – \Phi\!\Big(\frac{c – g\mu}{g}\Big), \qquad \text{false alarms} = 1 – \Phi\!\Big(\frac{c}{g}\Big), \qquad d' = \mu .

The last expression is the point of the chapter: discriminability does not depend on the gain. A gain amplifies signal and noise together, so it cancels in d′d'. Volume 34, Chapter 1 discussed this property of signal detection theory; it is used directly here.

Calculation

# Chapter 3. "Not interpreting the inner picture" as a gain term in a
# signal-detection model. An inner event (an image, a sensation) yields
# evidence e ~ N(mu, 1) when the situation is in fact threatening and N(0, 1)
# when it is not. Semantic elaboration is modelled as a gain g on the evidence
# AND a shift of the decision criterion; the non-judgmental instruction is g <
# 1. Exact Gaussian expressions, no simulation:
#   hit  = 1 - Phi((c - g mu) / g)      false alarm = 1 - Phi(c / g)
#   d'   = mu                           (gain cancels in d', which is the
#   whole point)
# So the gain changes WHICH errors are made, not how much information is
# there. Any claim that non-judgmental attention improves discrimination needs
# a different mechanism.
import math
from statistics import NormalDist

N = NormalDist()
MU = 1.2            # separation between threatening and benign inner events
# prior probability that the situation is in fact threatening
BASE_RATE = 0.15

def rates(g, c):
    hit = 1.0 - N.cdf((c - g * MU) / g)
    fa = 1.0 - N.cdf(c / g)
    return hit, fa

def cost(g, c, c_miss=4.0, c_fa=1.0):
    """Expected cost: a missed real threat costs c_miss, a false alarm costs
    c_fa."""

    hit, fa = rates(g, c)
    return BASE_RATE * (1 - hit) * c_miss + (1 - BASE_RATE) * fa * c_fa

def optimal_criterion(g, c_miss=4.0, c_fa=1.0):
    """Likelihood-ratio criterion that minimises expected cost, in evidence
    units."""
    lr = (1 - BASE_RATE) * c_fa / (BASE_RATE * c_miss)
    return g * (MU / 2.0 + math.log(lr) / MU)

if __name__ == "__main__":
    print(
        "d' = mu = {:.2f} for every gain g: the gain does not add"
        " information.".format(MU)
    )
    print(
        "\ngain g, criterion held at the g-matched optimum (miss:false alarm"
        " = 4:1)"
    )
    print("   g      criterion   hit rate   false alarms   expected cost")
    for g in (1.6, 1.3, 1.0, 0.7, 0.5, 0.3):
        c = optimal_criterion(g)
        h, f = rates(g, c)
        print("   {:4.1f} {:11.3f} {:10.3f} {:14.3f} {:15.4f}"
              .format(g, c, h, f, cost(g, c)))

    print("\nSame gains, criterion NOT re-tuned (left at the g = 1 optimum):")
    c1 = optimal_criterion(1.0)
    print("   g      criterion   hit rate   false alarms   expected cost")
    for g in (1.6, 1.3, 1.0, 0.7, 0.5, 0.3):
        h, f = rates(g, c1)
        print("   {:4.1f} {:11.3f} {:10.3f} {:14.3f} {:15.4f}"
              .format(g, c1, h, f, cost(g, c1)))

    print(
        "\nWith the criterion re-tuned the cost is identical at every gain"
        " ({:.4f}):".format(cost(1.0, optimal_criterion(1.0)))
    )
    print(
        "the gain is a pure rescaling. Without re-tuning, a LOWER gain"
        " moves the"
    )
    print(
        "operating point toward fewer false alarms and more misses, so the"
        " cost"
    )
    print(
        "rises when real threats are expensive. The instruction 'stop"
        " interpreting'"
    )
    print(
        "buys calm by trading away sensitivity to events that do matter;"
        " whether"
    )
    print(
        "that trade is worth making depends on the base rate and the costs,"
        " which"
    )
    print(
        "are exactly the quantities a practitioner cannot observe for"
        " themselves."
    )

    print(
        "\nbase rate at which re-tuned cost of g = 0.5 exceeds that of g ="
        " 1.0:"
    )
    print(
        "   (none: the costs are equal at every base rate once the"
        " criterion is"
    )
    print(
        "    re-tuned, which is the formal content of 'the gain adds no"
        " information')"
    )
# d' = mu = 1.20 for every gain g: the gain does not add information.
#
# gain g, criterion held at the g-matched optimum (miss:false alarm = 4:1)
#    g      criterion   hit rate   false alarms   expected cost
#     1.6       1.424      0.622          0.187          0.3857
#     1.3       1.157      0.622          0.187          0.3857
#     1.0       0.890      0.622          0.187          0.3857
#     0.7       0.623      0.622          0.187          0.3857
#     0.5       0.445      0.622          0.187          0.3857
#     0.3       0.267      0.622          0.187          0.3857
#
# Same gains, criterion NOT re-tuned (left at the g = 1 optimum):
#    g      criterion   hit rate   false alarms   expected cost
#     1.6       0.890      0.740          0.289          0.4016
#     1.3       0.890      0.697          0.247          0.3916
#     1.0       0.890      0.622          0.187          0.3857
#     0.7       0.890      0.471          0.102          0.4036
#     0.5       0.890      0.281          0.037          0.4634
#     0.3       0.890      0.039          0.002          0.5781
#
# With the criterion re-tuned the cost is identical at every gain (0.3857):
# the gain is a pure rescaling. Without re-tuning, a LOWER gain moves the
# operating point toward fewer false alarms and more misses, so the cost
# rises when real threats are expensive. The instruction 'stop interpreting'
# buys calm by trading away sensitivity to events that do matter; whether
# that trade is worth making depends on the base rate and the costs, which
# are exactly the quantities a practitioner cannot observe for themselves.
#
# base rate at which re-tuned cost of g = 0.5 exceeds that of g = 1.0:
#    (none: the costs are equal at every base rate once the criterion is
#     re-tuned, which is the formal content of 'the gain adds no information')

The contrast between the two tables says everything. With the criterion re-tuned to the gain, all six gains give identical hit rates (0.622), false-alarm rates (0.187) and expected costs (0.3857): the gain is a pure rescaling and changes nothing. Without re-tuning, as the gain falls from 1.6 to 0.3 the hit rate falls from 0.740 to 0.039 and false alarms from 0.289 to 0.002, while the expected cost (a miss costing 4, a false alarm 1) falls from 0.4016 to 0.3857 and then rises to 0.5781.

So the entire content of “non-judgmental” in this model is: with the criterion unchanged, move the operating point toward fewer false alarms and more misses. It adds no information. Whether that trade is worth making depends on the base rate and on the costs of the two kinds of error — which are exactly the two quantities a practitioner cannot observe for themselves.

Note

This does not say the practice is useless. If someone’s problem is precisely too many false alarms — reading harmless bodily sensations as danger — then moving the operating point toward fewer false alarms is the right thing to do, and that is what the cognitive model of panic says (Clark 1986): catastrophic interpretation of bodily sensation is what maintains the symptom. What this chapter says is that the benefit comes from the operating point having been wrong to begin with, not from “non-judgmental attention being more accurate”. Stated as the latter, it will be recommended to people whose operating point was not wrong, and for them it is pure loss.

Evidence

Affect labelling (human experiments and meta-analysis). The opposite practice — naming the emotion — also has support. Lieberman and colleagues reported reduced amygdala response and increased right ventrolateral prefrontal activity when participants attached emotion labels to facial expressions (Lieberman et al. 2007). Torre and Lieberman later wrote affect labelling as a form of implicit emotion regulation (Torre and Lieberman 2018).

This runs in the opposite direction to an instruction not to name, and both have their advocates. This book does not adjudicate the disagreement; it notes that it exists. Presenting “non-judgmental” as the only effective orientation is inconsistent with the affect-labelling literature, and the two bodies of work often measure different outcomes (neural response, self-reported intensity, behavioural avoidance).

Habituation and sensitisation: the time of measurement decides the answer

If someone repeatedly encounters the same kind of inner event, how does the response change? The standard answer is dual-process theory: habituation and sensitisation are independent processes and the net response is their difference (Groves and Thompson 1970). Rankin and colleagues give an updated review of the behavioural characteristics of habituation, including the influence of interstimulus interval and spontaneous recovery (Rankin et al. 2009).

Written as two first-order “stores”: a habituation store HH decaying slowly (over days) and a sensitisation store ZZ decaying quickly (over minutes), with response R=R0e−H(1+Z)R = R_0 e^{-H}(1+Z).

Calculation

# Chapter 3. Habituation and sensitisation as competing first-order processes
# (the standard dual-process account). Each exposure adds to a slowly decaying
# habituation store H and a quickly decaying sensitisation store Z:
#   response R_n = R0 * exp(-H_n) * (1 + Z_n)
#   between exposures separated by dt:  H <- H e^{-dt/tauH} + aH,  Z <- Z
#   e^{-dt/tauZ} + aZ
# tauH is long (days), tauZ short (under an hour). The program's point is NOT
# that spacing changes how much is learned -- in this model it barely does --
# but that the response measured AT THE END OF A SESSION is dominated by Z,
# while the response at a delayed test reflects H. So within-session change is
# a poor measure of learning, and two studies measuring at different times
# reach different answers from the same model. Author's parameterisation of a
# standard model; illustrative units.
import math

R0, A_H, A_Z = 1.0, 0.25, 0.35
TAU_H, TAU_Z = 10080.0, 40.0      # minutes: one week, versus forty minutes

def session(n_exp, dt_min):
    H = Z = 0.0
    R = []
    for _ in range(n_exp):
        R.append(R0 * math.exp(-H) * (1.0 + Z))
        H = H * math.exp(-dt_min / TAU_H) + A_H
        Z = Z * math.exp(-dt_min / TAU_Z) + A_Z
    return R, H, Z

def probe(H, Z, delay_min):
    Hd = H * math.exp(-delay_min / TAU_H)
    Zd = Z * math.exp(-delay_min / TAU_Z)
    return R0 * math.exp(-Hd) * (1.0 + Zd)

if __name__ == "__main__":
    print(
        "tauH = {:.0f} min ({:.1f} days), tauZ = {:.0f} min. Eight exposures."
        .format(TAU_H, TAU_H / 1440.0, TAU_Z)
    )
    print(
        "\n   interval  R1     R8     end of     H     Z    test at  test at"
    )
    print("                           session                 24 h     7 d")
    for dt in (2.0, 10.0, 30.0, 60.0, 120.0, 240.0, 1440.0):
        R, H, Z = session(8, dt)
        print(
            "   {:8.0f} {:6.3f} {:6.3f} {:10.3f} {:5.2f} {:5.2f} {:8.3f}"
            " {:8.3f}".format(
                dt,
                R[0],
                R[7],
                probe(H, Z, 0.0),
                H,
                Z,
                probe(H, Z, 1440.0),
                probe(H, Z, 10080.0),
            )
        )

    print(
        "\nRead the H column first: across a 720-fold range of intervals it"
        " moves"
    )
    print(
        "only between {:.2f} and {:.2f}, because habituation decays over days"
        " while".format(
            min(session(8, d)[1] for d in (2.0, 1440.0)),
            max(session(8, d)[1] for d in (2.0, 1440.0)),
        )
    )
    print(
        "the whole session is short by comparison. Z, by contrast, moves"
        " between"
    )
    print(
        "{:.2f} and {:.2f} over the same range. So in this model spacing"
        " changes the".format(
            min(session(8, d)[2] for d in (1440.0, 2.0)),
            max(session(8, d)[2] for d in (1440.0, 2.0)),
        )
    )
    print(
        "fast store far more than the slow one -- it changes what is MEASURED"
        " much"
    )
    print("more than what is learned:")
    for dt in (2.0, 240.0):
        R, H, Z = session(8, dt)
        print(
            "   at {:>4.0f}-minute intervals, Z at the end of the session is"
            " {:.2f} and the".format(dt, Z)
        )
        print("      end-of-session response is {:.3f}; at 24 h it is {:.3f}."
              .format(probe(H, Z, 0.0), probe(H, Z, 1440.0)))

    print(
        "\nThe methodological consequence is the one Chapter 8 acts on."
        " Measuring at"
    )
    print(
        "the end of a session confounds learning with a fast sensitisation"
        " process"
    )
    print(
        "that will be gone by the next day, and the sign of the confound"
        " depends on"
    )
    print(
        "the spacing. The protocol therefore takes its primary measurement"
        " at a"
    )
    print(
        "delayed test on a separate day, not at the end of a practice"
        " session."
    )

    print("\nWhat the within-session trace looks like, massed versus spaced:")
    for dt in (2.0, 240.0):
        R, _, _ = session(8, dt)
        print(
            "   {:>4.0f} min: ".format(dt)
            + " ".join("{:.3f}".format(v) for v in R)
        )
    print(
        "\n   The massed trace rises before it falls (sensitisation outruns"
    )
    print(
        "   habituation early on) while the spaced trace falls"
        " monotonically. A"
    )
    print(
        "   practitioner who stops because 'it got worse' stops during the"
        " rise."
    )

    print(
        "\nLimits of this model: both stores are first-order with fixed"
        " increments,"
    )
    print(
        "so it cannot represent saturation, context specificity, or the"
        " extinction"
    )
    print(
        "literature's distinction between losing a response and learning a"
        " new one."
    )
    print(
        "It is used here only to show that the TIME OF MEASUREMENT matters,"
        " which"
    )
    print("does not depend on those details.")
# tauH = 10080 min (7.0 days), tauZ = 40 min. Eight exposures.
#
#    interval  R1     R8     end of     H     Z    test at  test at
#                            session                 24 h     7 d
#           2  1.000  0.543      0.456  2.00  2.37    0.177    0.479
#          10  1.000  0.403      0.323  1.99  1.37    0.178    0.480
#          30  1.000  0.293      0.230  1.98  0.66    0.180    0.483
#          60  1.000  0.260      0.205  1.96  0.45    0.183    0.486
#         120  1.000  0.253      0.201  1.92  0.37    0.189    0.494
#         240  1.000  0.264      0.214  1.84  0.35    0.202    0.508
#        1440  1.000  0.412      0.376  1.28  0.35    0.330    0.625
#
# Read the H column first: across a 720-fold range of intervals it moves
# only between 1.28 and 2.00, because habituation decays over days while
# the whole session is short by comparison. Z, by contrast, moves between
# 0.35 and 2.37 over the same range. So in this model spacing changes the
# fast store far more than the slow one -- it changes what is MEASURED much
# more than what is learned:
#    at    2-minute intervals, Z at the end of the session is 2.37 and the
#       end-of-session response is 0.456; at 24 h it is 0.177.
#    at  240-minute intervals, Z at the end of the session is 0.35 and the
#       end-of-session response is 0.214; at 24 h it is 0.202.
#
# The methodological consequence is the one Chapter 8 acts on. Measuring at
# the end of a session confounds learning with a fast sensitisation process
# that will be gone by the next day, and the sign of the confound depends on
# the spacing. The protocol therefore takes its primary measurement at a
# delayed test on a separate day, not at the end of a practice session.
#
# What the within-session trace looks like, massed versus spaced:
#       2 min: 1.000 1.051 1.021 0.945 0.847 0.742 0.639 0.543
#     240 min: 1.000 1.051 0.824 0.649 0.515 0.410 0.328 0.264
#
#    The massed trace rises before it falls (sensitisation outruns
#    habituation early on) while the spaced trace falls monotonically. A
#    practitioner who stops because 'it got worse' stops during the rise.
#
# Limits of this model: both stores are first-order with fixed increments,
# so it cannot represent saturation, context specificity, or the extinction
# literature's distinction between losing a response and learning a new one.
# It is used here only to show that the TIME OF MEASUREMENT matters, which
# does not depend on those details.

The result differs from the usual account and is worth noting. With these parameters, the interval barely changes how much is learned: HH moves only from 1.28 to 2.00 across a 720-fold range of intervals. What the interval changes is the size of ZZ at the moment of measurement: at 2-minute intervals ZZ at the end of the session is 2.37 and the end-of-session response 0.456, falling to 0.177 a day later; at 240-minute intervals ZZ is 0.35, the end-of-session response 0.214 and the next-day value 0.202.

Two conclusions follow. The methodological one: measuring at the end of a session measures mainly a fast process that will be gone by the next day, and the direction of the confound depends on the spacing. Chapter 8 therefore places the primary measurement in an independent assessment on another day, not at the end of a practice session. The one for practitioners: the response curve of massed practice rises before it falls (1.000, 1.051, 1.021, 0.945 …) while more widely spaced practice falls monotonically. “It felt worse at first” is what this model predicts for massed practice: neither evidence that something has gone wrong nor evidence that it is working.

Note

The limits of this model. Both stores are first-order with fixed increments, so it cannot represent saturation, context specificity, or the extinction literature’s distinction between losing a response and learning a new inhibition (Craske et al. 2014). It is used here only to make the point about the time of measurement, and that point does not depend on those details.

Interoceptive exposure: clinicians have been doing this for decades

It is worth being clear that “deliberately evoking bodily sensations and experiencing them without avoidance” is not a new invention; in the treatment of panic disorder it is called interoceptive exposure.

Evidence

Interoceptive exposure (clinical review). Boettcher, Brake and Barlow review the origins and current state of interoceptive exposure: deliberately evoking palpitations, dizziness or breathlessness (by hyperventilation or spinning, for instance) so that the patient learns that the sensations themselves are not dangerous (Boettcher, Brake, and Barlow 2016). Its theoretical basis is Clark’s cognitive model of panic (Clark 1986). Craske and colleagues rewrote the mechanism of exposure therapy as inhibitory learning rather than habituation and derived a set of principles for optimising exposure from that (Craske et al. 2014).

This bears on the book’s framework in two ways. First, the activity the original claim describes overlaps heavily with an intervention that has clinical evidence, so “this class of practice may work” is not a bold claim. Second, the clinical version has a component the original claim lacks: it deliberately evokes sensations and remains with them, rather than “not interpreting sensations that arise”. On the inhibitory-learning account what does the work is the mismatch between expectation and outcome (Craske et al. 2014) — and that requires a definite expectation first, that is, it requires interpreting first and then being disconfirmed. That is not mechanistically compatible with “not interpreting”.

Can belief updating explain relapse

The inhibitory-learning framework poses a natural quantitative question: if the practice works by learning that “nothing follows this sensation”, how long does learning take? Written as Beta–Bernoulli updating, each uneventful occasion is one observation of 0.

Calculation

# Chapter 3. "Nothing bad will happen" as a Bayesian belief, and how many
# disconfirming trials it takes to move. A person holds P(catastrophe follows
# the sensation) with a Beta prior whose strength reflects how firmly the
# belief is held. Each exposure in which nothing happens is one Bernoulli
# observation of 0. Exact Beta-Bernoulli updating:
#   Beta(a, b) + k failures-to-occur -> Beta(a, b + k)
# The program reports how many clean trials are needed to bring the posterior
# mean below a threshold, for several prior strengths, and what ONE bad trial
# does. Standard library only; no simulation, exact arithmetic.
import math

def posterior(a, b, k_no_event, k_event=0):
    return a + k_event, b + k_no_event

def mean(a, b):
    return a / (a + b)

def trials_to(a, b, target):
    k = 0
    while mean(*posterior(a, b, k)) > target and k < 100000:
        k += 1
    return k

if __name__ == "__main__":
    print(
        "Prior belief that the sensation is followed by catastrophe, and the"
        " number"
    )
    print(
        "of clean exposures needed to bring the posterior mean under a"
        " threshold.\n"
    )
    print(
        "   prior Beta(a,b)   prior mean   to < 0.10   to < 0.05   to < 0.01"
    )
    priors = [(1.0, 1.0), (2.0, 1.0), (5.0, 1.0), (9.0, 1.0),
              (18.0, 2.0), (45.0, 5.0)]
    for a, b in priors:
        print("   Beta({:4.0f},{:3.0f}) {:13.3f} {:11d} {:11d} {:11d}"
              .format(a, b, mean(a, b), trials_to(a, b, 0.10),
                      trials_to(a, b, 0.05), trials_to(a, b, 0.01)))

    print(
        "\nA weakly held belief, Beta(2,1), is under 0.05 after {} clean"
        " trials.".format(trials_to(2.0, 1.0, 0.05))
    )
    print(
        "A firmly held one, Beta(45,5), needs {} - which at one exposure a"
        " day is".format(trials_to(45.0, 5.0, 0.05))
    )
    print(
        "about {:.1f} years. This is the quantitative content of 'it takes a"
        " long".format(trials_to(45.0, 5.0, 0.05) / 365.0)
    )
    print(
        "time': not a platitude, but a consequence of how much evidence a"
        " confident"
    )
    print("prior is worth.")

    print("\nAsymmetry: what ONE occurrence does, after 200 clean trials.")
    print(
        "   prior         after 200 clean   then 1 event   change   trials to"
        " undo"
    )
    for a, b in [(5.0, 1.0), (18.0, 2.0), (45.0, 5.0)]:
        a1, b1 = posterior(a, b, 200)
        m1 = mean(a1, b1)
        a2, b2 = posterior(a1, b1, 0, 1)
        m2 = mean(a2, b2)
        back = 0
        while mean(a2, b2 + back) > m1 and back < 100000:
            back += 1
        print("   Beta({:4.0f},{:3.0f}) {:16.4f} {:14.4f} {:+8.4f} {:15d}"
              .format(a, b, m1, m2, m2 - m1, back))

    print(
        "\nThis is where the model FAILS to reproduce the phenomenon it is"
        " offered"
    )
    print(
        "to explain. One occurrence after 200 clean trials moves the"
        " posterior mean"
    )
    print(
        "by between +0.003 and +0.005, and between 5 and 41 further clean"
        " trials"
    )
    print(
        "undo it. A Bayesian learner of this kind cannot relapse: by"
        " construction,"
    )
    print(
        "200 observations outweigh 1. Yet sudden return of fear after a"
        " single bad"
    )
    print(
        "episode is among the best-attested findings in the extinction"
        " literature."
    )
    print(
        "So the arithmetic above is evidence AGAINST simple evidential"
        " updating as"
    )
    print(
        "the mechanism, not evidence for it. Something else - context,"
        " attribution,"
    )
    print(
        "or a separate inhibitory memory rather than a revised probability -"
        " has to"
    )
    print("carry the asymmetry, and this model has no term for it.")

    print(
        "\nWhat the model does NOT say: it assumes the trials are"
        " independent and"
    )
    print(
        "that the person updates on them. Both fail in the cases that matter"
        " most -"
    )
    print(
        "avoidance makes trials non-independent (the person attributes safety"
        " to the"
    )
    print(
        "avoidance, not to the absence of danger), and a belief maintained"
        " for other"
    )
    print(
        "reasons will not update at all. Those are the standard reasons a"
        " purely"
    )
    print(
        "evidential account of exposure is incomplete; the arithmetic above"
        " is an"
    )
    print(
        "upper bound on how fast belief could move, not a prediction that it"
        " does."
    )
# Prior belief that the sensation is followed by catastrophe, and the number
# of clean exposures needed to bring the posterior mean under a threshold.
#
#    prior Beta(a,b)   prior mean   to < 0.10   to < 0.05   to < 0.01
#    Beta(   1,  1)         0.500           8          18          98
#    Beta(   2,  1)         0.667          17          37         197
#    Beta(   5,  1)         0.833          44          94         494
#    Beta(   9,  1)         0.900          80         170         890
#    Beta(  18,  2)         0.900         160         340        1780
#    Beta(  45,  5)         0.900         400         850        4450
#
# A weakly held belief, Beta(2,1), is under 0.05 after 37 clean trials.
# A firmly held one, Beta(45,5), needs 850 - which at one exposure a day is
# about 2.3 years. This is the quantitative content of 'it takes a long
# time': not a platitude, but a consequence of how much evidence a confident
# prior is worth.
#
# Asymmetry: what ONE occurrence does, after 200 clean trials.
#    prior         after 200 clean   then 1 event   change   trials to undo
#    Beta(   5,  1)           0.0243         0.0290  +0.0047              41
#    Beta(  18,  2)           0.0818         0.0860  +0.0042              12
#    Beta(  45,  5)           0.1800         0.1833  +0.0033               5
#
# This is where the model FAILS to reproduce the phenomenon it is offered
# to explain. One occurrence after 200 clean trials moves the posterior mean
# by between +0.003 and +0.005, and between 5 and 41 further clean trials
# undo it. A Bayesian learner of this kind cannot relapse: by construction,
# 200 observations outweigh 1. Yet sudden return of fear after a single bad
# episode is among the best-attested findings in the extinction literature.
# So the arithmetic above is evidence AGAINST simple evidential updating as
# the mechanism, not evidence for it. Something else - context, attribution,
# or a separate inhibitory memory rather than a revised probability - has to
# carry the asymmetry, and this model has no term for it.
#
# What the model does NOT say: it assumes the trials are independent and
# that the person updates on them. Both fail in the cases that matter most -
# avoidance makes trials non-independent (the person attributes safety to the
# avoidance, not to the absence of danger), and a belief maintained for other
# reasons will not update at all. Those are the standard reasons a purely
# evidential account of exposure is incomplete; the arithmetic above is an
# upper bound on how fast belief could move, not a prediction that it does.

The first part answers “how long” quantitatively: a weakly held belief, Beta(2,1), falls below 0.05 after 37 clean trials, while a firmly held Beta(45,5) needs 850, about 2.3 years at one exposure a day. That turns “it takes a long time” from a platitude into a number.

The second part is the chapter’s second refutation. After 200 clean trials, one event raises the posterior mean by only 0.0033 to 0.0047, and 5 to 41 clean trials undo it. That is, a Bayesian learner of this kind cannot relapse — by construction, 200 observations outweigh 1. Yet the rapid return of fear after a single bad episode is among the best-attested findings in the extinction literature (Bouton 2002). So this arithmetic is evidence against simple evidential updating as the mechanism. Something else must carry the asymmetry: context, attribution, or a separate inhibitory memory rather than a revised probability (Bouton 2002; Craske et al. 2014).

How to measure a stream that is not being interpreted

A last methodological problem. If the practice requires not interpreting inner pictures, how is a researcher to measure that stream of pictures without asking the participant to interpret it? One route that demands no semantic judgement is to record only changes of category, and then compute the transition entropy of that categorical sequence: high entropy for a flowing, non-repeating stream, low entropy for a stream that keeps returning to the same category.

The difficulty here is statistical. Entropy estimates on short sequences carry a systematic downward bias; Miller gave the first-order correction long ago (Miller 1955) and Paninski analysed the properties of entropy and mutual information estimators systematically (Paninski 2003).

Calculation

# Chapter 3. Measuring an imagery stream without interpreting it: transition
# entropy of a categorical sequence, and why the estimate is biased when
# sequences are short. A practitioner reports the category of each inner event
# at fixed intervals (K = 6 coarse categories). Two generators: a FREE stream
# (near-uniform transitions) and a STUCK stream (high probability of repeating
# the same category, the formal version of "going round the same thought").
# The plug-in entropy estimate is biased DOWNWARD, and the bias is of the same
# size as the effect being looked for, so the program applies the Miller-Madow
# correction and reports both. Synthetic data; fixed seed; standard library.
import math
import random

K = 6

def chain(stickiness):
    """Transition matrix with P(stay) = stickiness, rest spread uniformly."""
    off = (1.0 - stickiness) / (K - 1)
    return [
        [stickiness if i == j else off for j in range(K)] for i in range(K)
    ]

def sample(P, n, rng):
    s = rng.randrange(K)
    out = [s]
    for _ in range(n - 1):
        r, acc = rng.random(), 0.0
        for j in range(K):
            acc += P[s][j]
            if r <= acc:
                s = j
                break
        out.append(s)
    return out

def true_entropy_rate(P):
    """Conditional entropy H(X_{t+1} | X_t). The chain is symmetric, so its
    stationary
    distribution is uniform and every row has the same entropy: one row
    suffices.
    """
    return -math.fsum(p * math.log(p, 2) for p in P[0] if p > 0)

def plugin_rate(seq):
    """Plug-in conditional entropy H(X_{t+1} | X_t) in bits."""
    counts = [[0] * K for _ in range(K)]
    for a, b in zip(seq, seq[1:]):
        counts[a][b] += 1
    total = len(seq) - 1
    h = 0.0
    for i in range(K):
        ri = sum(counts[i])
        if ri == 0:
            continue
        for j in range(K):
            if counts[i][j]:
                p = counts[i][j] / ri
                h -= (ri / total) * p * math.log(p, 2)
    return h

def miller_madow(seq):
    """Plug-in plus (m - 1) / (2 N ln 2), m = number of observed transition
    cells."""
    counts = [[0] * K for _ in range(K)]
    for a, b in zip(seq, seq[1:]):
        counts[a][b] += 1
    total = len(seq) - 1
    nonzero_rows = sum(1 for i in range(K) if sum(counts[i]) > 0)
    m = sum(1 for i in range(K) for j in range(K) if counts[i][j] > 0)
    return plugin_rate(seq) + (m - nonzero_rows) / (2.0 * total * math.log(2))

if __name__ == "__main__":
    rng = random.Random(606)
    print(
        "K = {} categories. True conditional entropy from the generator,"
        " against".format(K)
    )
    print("estimates from sequences of the length a real session provides.\n")
    print("                             n=40          n=200         n=2000")
    print(
        "   stream  P(stay)  true H  plug-in  MM    plug-in  MM   "
        " plug-in  MM"
    )
    for lab, stay in (("free ", 1.0 / K), ("mild ", 0.35), ("stuck", 0.70)):
        P = chain(stay)
        th = true_entropy_rate(P)
        cells = []
        for n in (40, 200, 2000):
            pl = [plugin_rate(sample(P, n, rng)) for _ in range(300)]
            mm = [miller_madow(sample(P, n, rng)) for _ in range(300)]
            cells.append((math.fsum(pl) / len(pl), math.fsum(mm) / len(mm)))
        print(
            "   {}  {:7.3f} {:7.3f} {:8.3f} {:6.3f} {:8.3f} {:6.3f}"
            " {:8.3f} {:6.3f}".format(
                lab,
                stay,
                th,
                cells[0][0],
                cells[0][1],
                cells[1][0],
                cells[1][1],
                cells[2][0],
                cells[2][1],
            )
        )

    Pf, Pm = chain(1.0 / K), chain(0.35)
    hf, hm = true_entropy_rate(Pf), true_entropy_rate(Pm)
    pl40 = (
        math.fsum(plugin_rate(sample(Pf, 40, rng)) for _ in range(300)) / 300
    )
    mm40 = (
        math.fsum(miller_madow(sample(Pf, 40, rng)) for _ in range(300)) / 300
    )
    print(
        "\nAt n = 40, the length of a 20-minute session sampled every 30"
        " s, the"
    )
    print(
        "plug-in estimate understates the free stream by {:.2f} bits, and the"
        .format(hf - pl40)
    )
    print(
        "Miller-Madow correction cuts that to {:.2f} bits. For comparison,"
        " the whole".format(hf - mm40)
    )
    print(
        "true difference between the free and the mild stream is {:.2f}"
        " bits: the".format(hf - hm)
    )
    print(
        "uncorrected bias is about {:.0f} times the effect being looked"
        " for, and".format((hf - pl40) / (hf - hm))
    )
    print(
        "even the corrected bias is about {:.0f} times it. An uncorrected"
        " estimator".format((hf - mm40) / (hf - hm))
    )
    print(
        "would report differences in sequence LENGTH as differences in the"
        " stream."
    )

    print(
        "\nCan a session-length sequence tell the two streams apart?"
        " Distribution of"
    )
    print("the Miller-Madow estimate, 2000 sessions of each, n = 40:")
    for lab, stay in (("free ", 1.0 / K), ("mild ", 0.35)):
        P = chain(stay)
        v = sorted(miller_madow(sample(P, 40, rng)) for _ in range(2000))
        print("   {}: median {:.3f}, 5th-95th percentile {:.3f} to {:.3f}"
              .format(lab, v[1000], v[100], v[1900]))
    vf = sorted(miller_madow(sample(Pf, 40, rng)) for _ in range(2000))
    vm = sorted(miller_madow(sample(Pm, 40, rng)) for _ in range(2000))
    import bisect
    # P(a random mild session scores below a random free session) = the AUC
    auc = math.fsum(bisect.bisect_left(vm, x) for x in vf) / (
        len(vf) * len(vm)
    )
    print("   P(a mild session scores below a free one) = {:.3f}".format(auc))
    print("   (0.5 would be no discrimination at all, 1.0 perfect)")
    n_needed = None
    for n_sessions in range(1, 61):
        # SE of the per-person mean over n sessions, pooled SD from the two
        # arms
        sd = math.sqrt(
            0.5
            * (
                sum((x - vf[1000]) ** 2 for x in vf) / len(vf)
                + sum((x - vm[1000]) ** 2 for x in vm) / len(vm)
            )
        )
        if abs(vf[1000] - vm[1000]) / (sd / math.sqrt(n_sessions)) >= 2.8:
            n_needed = n_sessions
            break
    print(
        "\n   A single session discriminates at {:.2f}, which is far too"
        " weak to".format(auc)
    )
    print(
        "   classify one person's stream. Averaging over sessions per person,"
        " about"
    )
    print(
        "   {} sessions are needed before the person-level means of the two"
        " streams".format(n_needed)
    )
    print(
        "   are separated by 2.8 standard errors. That is why the Chapter 8"
        " protocol"
    )
    print(
        "   pre-specifies a per-person average over sessions and the"
        " Miller-Madow"
    )
    print(
        "   estimator, and records the sampling interval, which fixes n and"
        " so fixes"
    )
    print(
        "   the bias: an arm that reports more events per session would"
        " otherwise"
    )
    print("   score as having a higher-entropy stream for that reason alone.")
# K = 6 categories. True conditional entropy from the generator, against
# estimates from sequences of the length a real session provides.
#
#                              n=40          n=200         n=2000
#    stream  P(stay)  true H  plug-in  MM    plug-in  MM    plug-in  MM
#    free     0.167   2.585    1.928  2.265    2.472  2.579    2.574  2.585
#    mild     0.350   2.443    1.793  2.093    2.329  2.438    2.433  2.444
#    stuck    0.700   1.578    1.062  1.217    1.456  1.555    1.571  1.578
#
# At n = 40, the length of a 20-minute session sampled every 30 s, the
# plug-in estimate understates the free stream by 0.65 bits, and the
# Miller-Madow correction cuts that to 0.34 bits. For comparison, the whole
# true difference between the free and the mild stream is 0.14 bits: the
# uncorrected bias is about 5 times the effect being looked for, and
# even the corrected bias is about 2 times it. An uncorrected estimator
# would report differences in sequence LENGTH as differences in the stream.
#
# Can a session-length sequence tell the two streams apart? Distribution of
# the Miller-Madow estimate, 2000 sessions of each, n = 40:
#    free : median 2.267, 5th-95th percentile 2.005 to 2.500
#    mild : median 2.108, 5th-95th percentile 1.785 to 2.396
#    P(a mild session scores below a free one) = 0.741
#    (0.5 would be no discrimination at all, 1.0 perfect)
#
#    A single session discriminates at 0.74, which is far too weak to
#    classify one person's stream. Averaging over sessions per person, about
#    9 sessions are needed before the person-level means of the two streams
#    are separated by 2.8 standard errors. That is why the Chapter 8 protocol
#    pre-specifies a per-person average over sessions and the Miller-Madow
#    estimator, and records the sampling interval, which fixes n and so fixes
#    the bias: an arm that reports more events per session would otherwise
#    score as having a higher-entropy stream for that reason alone.

The magnitudes are worth recording: at n=40n = 40 (twenty minutes sampled every thirty seconds) the naive plug-in estimate understates the free stream by 0.65 bits and the Miller–Madow correction cuts that to 0.34 bits, while the true difference between the free and the mildly sticky stream is only 0.14 bits. The uncorrected bias is about five times the effect sought, and the corrected bias about twice it. More importantly, the bias varies with sequence length, so an arm reporting more events would be estimated as having a higher-entropy stream for that reason alone.

A single session discriminates at only 0.74 (chance being 0.5), far too weakly to classify one person’s stream; averaging over sessions per person, about nine sessions are needed before the person-level means of the two streams are separated by 2.8 standard errors. Chapter 8 accordingly specifies a per-person average over sessions, the Miller–Madow estimator, and a recorded sampling interval.

Hypothesis · not established

The author’s hypothesis (not established): a practice aimed at not interpreting pictures will raise the transition entropy (the flowingness) of the self-reported imagery stream, and that change will precede the fall in the body-vigilance score. What would show it to be wrong: in a pre-registered trial, finding that the 95 % confidence interval for the week-2 difference in imagery-stream transition entropy (Miller–Madow, per-person mean, fixed sampling interval) between the practice arm and the structurally matched control includes zero, or that the change does not precede the change in the primary outcome, would refute it.

Note

Not everyone has imagery. The vividness of visual imagery varies widely, and some people have no spontaneous visual imagery at all — congenital aphantasia (Zeman, Dewar, and Della Sala 2015); Pearson reviews the neuroscience of imagery, including individual differences in its strength and their measurement (Pearson 2019), and Dijkstra and colleagues review the neural mechanisms shared by imagery and perception (Dijkstra, Bosch, and Gerven 2019). A practice built around “watching an inner picture” does not apply to people without imagery; any material promoting such practices should say so, and Chapter 8 makes imagery vividness a pre-specified stratification variable and tests it as an effect modifier.

Summary

Written as a gain term in a signal-detection model, “not interpreting the inner picture” leaves discriminability d′d' untouched: with the criterion re-tuned, the expected cost is identical at every gain (0.3857); without re-tuning, a gain of 0.5 cuts the hit rate from 0.622 to 0.281 and raises the cost from 0.3857 to 0.4634. Non-judgmental attention trades one class of error for another rather than adding information; its benefit comes from the operating point having been wrong to begin with, which is exactly the situation described by the cognitive model of panic. The affect-labelling literature offers an equally supported practice running in the opposite direction. The dual-process model shows that spacing barely changes how much is learned but changes the fast process measured at the end of a session, so the primary measurement belongs on another day. Beta–Bernoulli updating supplies a number for “how long” (about 850 clean trials for a firmly held belief) and simultaneously fails to explain relapse, making it evidence against simple evidential updating. Finally, transition entropy on session-length sequences carries a bias two to five times the effect sought, and must be corrected, with a fixed sampling interval and per-person averaging.

References

Bishop, Scott R., Mark Lau, Shauna Shapiro, Linda Carlson, Nicole D. Anderson, James Carmody, Zindel V. Segal, et al. 2004. “Mindfulness: A Proposed Operational Definition.” Clinical Psychology: Science and Practice 11 (3): 230–41. https://doi.org/10.1093/clipsy.bph077.
Boettcher, Hannah, C. Alex Brake, and David H. Barlow. 2016. “Origins and Outlook of Interoceptive Exposure.” Journal of Behavior Therapy and Experimental Psychiatry 53: 41–51. https://doi.org/10.1016/j.jbtep.2015.10.009.
Bouton, Mark E. 2002. “Context, Ambiguity, and Unlearning: Sources of Relapse After Behavioral Extinction.” Biological Psychiatry 52 (10): 976–86. https://doi.org/10.1016/S0006-3223(02)01546-9.
Clark, David M. 1986. “A Cognitive Approach to Panic.” Behaviour Research and Therapy 24 (4): 461–70. https://doi.org/10.1016/0005-7967(86)90011-2.
Craske, Michelle G., Michael Treanor, Christopher C. Conway, Tomislav Zbozinek, and Bram Vervliet. 2014. “Maximizing Exposure Therapy: An Inhibitory Learning Approach.” Behaviour Research and Therapy 58: 10–23. https://doi.org/10.1016/j.brat.2014.04.006.
Dijkstra, Nadine, Sander E. Bosch, and Marcel A. J. van Gerven. 2019. “Shared Neural Mechanisms of Visual Perception and Imagery.” Trends in Cognitive Sciences 23 (5): 423–34. https://doi.org/10.1016/j.tics.2019.02.004.
Groves, Philip M., and Richard F. Thompson. 1970. “Habituation: A Dual-Process Theory.” Psychological Review 77 (5): 419–50. https://doi.org/10.1037/h0029810.
Hölzel, Britta K., Sara W. Lazar, Tim Gard, Zev Schuman-Olivier, David R. Vago, and Ulrich Ott. 2011. “How Does Mindfulness Meditation Work? Proposing Mechanisms of Action from a Conceptual and Neural Perspective.” Perspectives on Psychological Science 6 (6): 537–59. https://doi.org/10.1177/1745691611419671.
Lieberman, Matthew D., Naomi I. Eisenberger, Molly J. Crockett, Sabrina M. Tom, Jennifer H. Pfeifer, and Baldwin M. Way. 2007. “Putting Feelings into Words: Affect Labeling Disrupts Amygdala Activity in Response to Affective Stimuli.” Psychological Science 18 (5): 421–28. https://doi.org/10.1111/j.1467-9280.2007.01916.x.
Miller, George A. 1955. “Note on the Bias of Information Estimates.” In Information Theory in Psychology: Problems and Methods, edited by Henry Quastler, 95–100. Glencoe, Illinois: Free Press.
Paninski, Liam. 2003. “Estimation of Entropy and Mutual Information.” Neural Computation 15 (6): 1191–1253. https://doi.org/10.1162/089976603321780272.
Pearson, Joel. 2019. “The Human Imagination: The Cognitive Neuroscience of Visual Mental Imagery.” Nature Reviews Neuroscience 20: 624–34. https://doi.org/10.1038/s41583-019-0202-9.
Rankin, Catharine H., Thomas Abrams, Robert J. Barry, Seema Bhatnagar, David F. Clayton, John Colombo, Gianluca Coppola, et al. 2009. “Habituation Revisited: An Updated and Revised Description of the Behavioral Characteristics of Habituation.” Neurobiology of Learning and Memory 92 (2): 135–38. https://doi.org/10.1016/j.nlm.2008.09.012.
Torre, Jared B., and Matthew D. Lieberman. 2018. “Putting Feelings into Words: Affect Labeling as Implicit Emotion Regulation.” Emotion Review 10 (2): 116–24. https://doi.org/10.1177/1754073917742706.
Zeman, Adam, Michaela Dewar, and Sergio Della Sala. 2015. “Lives Without Imagery — Congenital Aphantasia.” Cortex 73: 378–80. https://doi.org/10.1016/j.cortex.2015.05.019.

Fan, Y. (2026). Interoceptive Decoupling, Salience-Network Switching and Autonomic Visceral Restoration: From Precision-Weighting in the Default Mode Network and the Redistribution of Vagal Tone to the Neurophenomenology of a Non-Judgmental Imagery Stream, with a Textual Examination of neiguan in the Huangdi Neijing [内感受解离、突显网络切换与内脏自主神经修复] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24020418 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).