5D Library › Two-Scale Renormalization, Occupancy Entropy and Thrust-Wave Superposition in Discrete Cyclical Networks · 中文

Chapter 4 The Zero-Thrust State and Felt Resistance: The Weber–Fechner Law, Adaptation Level and Habituation (Analogy and Hypothesis)

Note

What this chapter is. The cells of the first three chapters feel nothing. This chapter borrows concepts from psychophysics to ask an analogical question: if a cell’s excitation FF is treated as the “stimulus” received by some subject, how would the frequent zero excitations of “Wave-Rider” mode (the author’s “zero-thrust vacuum state”) be felt? This is an analogy and a hypothesis that carries psychophysics over to model variables, not a known psychological fact; the “subject” and the numbers are illustrative. The chapter does not concern the diagnosis of, or advice about, any mental-health problem.

The Weber–Fechner law: a review

5D Library volume 34 treated the Weber–Fechner law systematically; here we recall only three points that matter for this chapter. First, Weber found that the just-noticeable increment in a stimulus is roughly proportional to the stimulus’s intensity; Fechner assumed on this basis that every just-noticeable difference is equal in sensation, and integration gave sensation proportional to the logarithm of stimulus intensity, S=kln⁡(I/I0)S = k \ln (I / I_0), where I0I_0 is the absolute threshold (Fechner 1860). Second, using magnitude estimation, Stevens found that many sensory channels are better described by a power function S=kIβS = k I^{\beta}, with an exponent that differs from channel to channel (Stevens 1957). Third, the logarithmic law has a boundary that bears directly on this chapter: below the threshold I0I_0 Fechner’s formula gives negative values, while what actually occurs is “no sensation”, not “negative sensation”. The Weber–Fechner law on its own therefore does not imply that zero stimulation is felt as strong resistance; another mechanism is needed.

Adaptation level and the reference point

Evidence

Adaptation-level theory. Helson proposed in 1947 that when people judge the magnitude of a stimulus, their reference is an “adaptation level”, a weighted average of recently received stimuli (together with background and residual stimulation). Stimuli above the adaptation level are judged “heavy”, “bright” or “good”, those below it “light”, “dark” or “bad”, and a stimulus at the adaptation level is judged neutral (Helson 1947). His 1964 monograph extended the idea to many domains of judgement (Helson 1964). Kahneman and Tversky’s prospect theory of 1979 likewise puts a reference point at its centre: people evaluate an outcome according to whether it is a gain or a loss relative to the reference point, and losses are felt more strongly than gains of the same size (Kahneman and Tversky 1979).

Evidence

The behavioural characteristics of habituation. Thompson and Spencer summarised in 1966 a set of behavioural characteristics of habituation: when the same stimulus is presented repeatedly, the response gradually weakens; when stimulation stops for a while, the response recovers spontaneously; the more frequent the stimulation, the faster the habituation; and after another strong stimulus, a weakened response can be restored (dishabituation) (Thompson and Spencer 1966). In 2009 Rankin and thirteen co-authors re-examined these characteristics, changing a few substantively and adding one about long-term habituation (Rankin et al. 2009). Note that habituation concerns the weakening of a response to repeated stimulation; it is not the same concept as sensory adaptation, nor as Helson’s adaptation level. This chapter borrows only its form: the response weakens with repetition and recovers when the stimulus stops.

The “habituation-baseline deprivation” of the author’s framework can be read as follows. A subject in the “superposed” state receives strong excitation over and over; its adaptation level rises and its response to strong excitation weakens. Once it moves into the “zero-thrust” state, the excitation it receives lies far below the raised adaptation level and is felt as a large negative gap. The following model makes this reading explicit.

Calculation

The author’s model (illustrative parameters). The subject sits on one cell of the ring and at step tt receives excitation FtF_t (computed as in Chapter 3). The thrust it feels is a logarithmic contrast against an adaptation level AtA_t: ψt=ln⁡Ft+εAt+ε,At+1=At+λ(Ft−At),\psi_t = \ln \frac{F_t + \varepsilon}{A_t + \varepsilon},\qquad A_{t+1} = A_t + \lambda\,(F_t – A_t), where λ\lambda is the speed of adaptation and ε>0\varepsilon > 0 is a floor (something like a threshold) that keeps the logarithm defined at zero excitation. Two state sequences are compared: in “Fixed-Mode” every step is “ten 1s and two 2s” placed at random; in “Wave-Rider” every step throws the 14 particles at random into 4 randomly chosen cells. A step with ψt<−1\psi_t < -1 counts as “strong resistance”. The final part gives a habituation-like time course and the effect of the floor ε\varepsilon.

Calculation

# Chapter 4 ANALOGY as a model (illustrative parameters): one node feels its excitation F
# as psi = ln((F + e) / (A + e)) against an adaptation level A <- A + lam (F - A).
# fixed: ten 1s and two 2s placed at random; wave: 14 particles into 4 random cells.
import math, random
N, K, C, E = 14, 12, 0.5, 0.25

def node_excitation(n, i=0):
    return n[i] if n[i] else C * n[(i + 6) % K]

def fixed_state(rng):
    n = [1] * K
    for j in rng.sample(range(K), 2): n[j] = 2
    return n

def wave_state(rng):
    cells = rng.sample(range(K), 4); n = [0] * K
    for _ in range(N): n[rng.choice(cells)] += 1
    return n

def run(make, lam, steps, rng, e=E):
    fs = [node_excitation(make(rng)) for _ in range(steps)]
    a = sum(fs) / len(fs); psis = []
    for f in fs:
        psis.append(math.log((f + e) / (a + e))); a += lam * (f - a)
    return fs, psis

if __name__ == "__main__":
    rng = random.Random(1860)
    print("regime  lam   mean F  SD F   P(F=0)  mean psi  P(psi < -1)  mean psi | F=0")
    for name, make in (("fixed", fixed_state), ("wave", wave_state)):
        for lam in (0.05, 0.2, 0.5):
            fs, ps = run(make, lam, 20000, rng)
            mf = sum(fs) / len(fs)
            sd = math.sqrt(sum((f - mf) ** 2 for f in fs) / len(fs))
            zero = [p for f, p in zip(fs, ps) if f == 0]
            mz = f"{sum(zero) / len(zero):+.2f}" if zero else "   -"
            neg = sum(p < -1 for p in ps) / len(ps)
            print(f"{name:6s}  {lam:.2f}  {mf:5.2f}  {sd:5.2f}  {len(zero) / len(fs):.3f}"
                  f"    {sum(ps) / len(ps):+.3f}     {neg:.3f}        {mz}")
    # Habituation-like time course: baseline 1, then a run of 2s, then withdrawal to 0.
    lam, a = 0.2, 1.0; course = [1] * 5 + [2] * 15 + [0] * 10 + [1] * 10; out = []
    for f in course:
        out.append(math.log((f + E) / (a + E))); a += lam * (f - a)
    print("time course, lam 0.2: run of 2s, steps 6, 10, 20 ->",
          ", ".join(f"{out[t - 1]:+.2f}" for t in (6, 10, 20)))
    print("   zeros, steps 21, 25, 30 ->",
          ", ".join(f"{out[t - 1]:+.2f}" for t in (21, 25, 30)),
          "| back to 1, step 31 ->", f"{out[30]:+.2f}")
    # The floor e (an arbitrary "threshold") drives much of the result for F = 0.
    for e in (0.1, 0.25, 1.0):
        fs, ps = run(wave_state, 0.2, 20000, rng, e)
        print(f"wave, lam 0.2, floor e = {e:4.2f}: mean psi {sum(ps) / len(ps):+.3f},"
              f" P(psi < -1) {sum(p < -1 for p in ps) / len(ps):.3f}")
# regime  lam   mean F  SD F   P(F=0)  mean psi  P(psi < -1)  mean psi | F=0
# fixed   0.05   1.17   0.37  0.000    -0.026     0.000           -
# fixed   0.20   1.16   0.37  0.000    -0.023     0.000           -
# fixed   0.50   1.17   0.37  0.000    -0.017     0.000           -
# wave    0.05   1.59   1.81  0.431    -0.638     0.436        -1.98
# wave    0.20   1.58   1.81  0.434    -0.595     0.438        -1.94
# wave    0.50   1.59   1.81  0.430    -0.464     0.389        -1.80
# time course, lam 0.2: run of 2s, steps 6, 10, 20 -> +0.59, +0.20, +0.02
#    zeros, steps 21, 25, 30 -> -2.18, -1.44, -0.72 | back to 1, step 31 -> +1.00
# wave, lam 0.2, floor e = 0.10: mean psi -0.937, P(psi < -1) 0.446
# wave, lam 0.2, floor e = 0.25: mean psi -0.595, P(psi < -1) 0.435
# wave, lam 0.2, floor e = 1.00: mean psi -0.215, P(psi < -1) 0.172

How to read the results

The subject’s mean excitation differs between the two modes: about 1.17 in Fixed-Mode and about 1.59 in Wave-Rider. The latter is higher, because empty cells borrow half the excitation of the cell opposite (Chapter 3). But Wave-Rider excitation fluctuates strongly (standard deviation 1.81 against 0.37), and about 43 % of the time it is zero. The outcome, at λ=0.2\lambda = 0.2: in Fixed-Mode the mean felt thrust is close to 0 (−0.023) and strong resistance never occurs; in Wave-Rider the mean felt thrust is −0.595, about 44 % of steps are strong resistance, and the mean feeling at zero excitation is −1.94. Varying the speed of adaptation λ\lambda from 0.05 to 0.5 changes these figures only moderately.

The time course shows the model’s form of “habituation” and “deprivation”. After a switch from a baseline of 1 to a run of 2s, felt thrust falls from +0.59 to +0.20 and +0.02: the subject gradually gets used to the sustained strong stimulus. When excitation then drops suddenly to zero, the first step is felt as −2.18, weakening to −1.44 and −0.72 as the adaptation level falls. On the return to 1 the feeling is +1.00, far above the original baseline.

This result, however, comes mainly from two mathematical choices in the model, not from psychology. First, the logarithm is concave, so by Jensen’s inequality the mean logarithm of a strongly fluctuating excitation sequence must lie below the logarithm of its mean; even with a fixed adaptation level, Wave-Rider’s mean feeling would be negative. Second, the feeling at zero excitation is set entirely by the arbitrary floor ε\varepsilon: with ε=0.1\varepsilon = 0.1 the mean feeling is −0.937, with ε=1.0\varepsilon = 1.0 only −0.215, and the share of strong-resistance steps falls from 44.6 % to 17.2 %. “The zero-thrust state necessarily brings strong felt resistance” is thus not a robust conclusion of the model but a consequence of parameter choices. The framework’s statement should be qualified to: under the two assumptions of logarithmic contrast and a low floor, a fluctuating excitation sequence with many zeros may be felt as worse on average, even when its mean is higher.

Hypothesis · not established

Author’s hypothesis: frequent-zero reward sequences and momentary feeling (not yet established). The author conjectures that when mean reward is equal (or slightly higher in the Wave-Rider group), the group whose reward sequence has frequent zeros and large fluctuations reports a lower average momentary feeling than the group with steady rewards, and that the difference is largest just after a zero reward. A testable prediction: in a pre-registered experiment, participants do the same simple task and are randomised to a “steady” or a “Wave-Rider” scoring sequence (the Wave-Rider mean being no lower), and after each round rate their momentary feeling on the same scale; the Wave-Rider group’s mean momentary feeling is lower, with a 95 % confidence interval for the difference that excludes 0. If the interval includes 0 and excludes a smallest effect of interest fixed in advance, or the difference goes the other way, the hypothesis is refuted. The experiment should use only token points or a small fixed payment, never a real investment or gambling setting.

Note

Division of labour with volume 34. Volume 34 discussed the measurement of Weber fractions, the assumptions of Fechner’s integration and the dispute over Stevens’s power law; this chapter does not repeat the programs for them. The model here uses only the simplest logarithmic contrast. With a power function or a piecewise function with a threshold, the feeling at zero excitation would be quite different, which is exactly the point stressed above.

Chapter summary

The Weber–Fechner law by itself says only that sensation is related to the logarithm of the stimulus and gives no “negative sensation” below threshold, so it does not imply that the zero-thrust state brings felt resistance. Helson’s adaptation level and the reference point of prospect theory supply the missing mechanism: feeling depends on the gap between a stimulus and the recent level. Research on habituation shows that responses to repeated stimuli weaken and recover when the stimulus stops. Written as an illustrative model, Wave-Rider gives higher mean excitation than Fixed-Mode (1.59 against 1.17) yet a lower mean feeling (−0.595 against −0.023), with about 44 % of steps at “strong resistance”; but the result comes mainly from the concavity of the logarithm and an arbitrary floor, and raising the floor from 0.1 to 1.0 moves the mean feeling from −0.937 to −0.215. The next chapter discusses the author’s four cross-domain applications and asks what is missing before a formal resemblance becomes a testable hypothesis.

References

Fechner, Gustav Theodor. 1860. Elemente der Psychophysik. Breitkopf und Härtel.
Helson, Harry. 1947. “Adaptation-Level as Frame of Reference for Prediction of Psychophysical Data.” American Journal of Psychology 60 (1): 1–29. https://doi.org/10.2307/1417326.
Helson, Harry. 1964. Adaptation-Level Theory: An Experimental and Systematic Approach to Behavior. Harper; Row.
Kahneman, Daniel, and Amos Tversky. 1979. “Prospect Theory: An Analysis of Decision Under Risk.” Econometrica 47 (2): 263–91. https://doi.org/10.2307/1914185.
Rankin, Catharine H., Thomas Abrams, Robert J. Barry, 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.
Stevens, S. S. 1957. “On the Psychophysical Law.” Psychological Review 64 (3): 153–81. https://doi.org/10.1037/h0046162.
Thompson, Richard F., and William A. Spencer. 1966. “Habituation: A Model Phenomenon for the Study of Neuronal Substrates of Behavior.” Psychological Review 73 (1): 16–43. https://doi.org/10.1037/h0022681.

Fan, Y. (2026). Two-Scale Renormalization, Occupancy Entropy and Thrust-Wave Superposition in Discrete Cyclical Networks: Algebraic Topology and Monte Carlo Verification on a Twelve-Node Ring Lattice [离散周期网络中的双尺度重整化、占位数分布熵与推力波叠加定理] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23796761 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).