Hysteresis¶

Prime #: 48
Origin domain: Physics
Also from: Economics & Finance, Biology & Ecology
Related primes: Irreversibility, Half-Life, Oscillation, Damping, Frame of Reference, State and State Transition, Tipping Points (or Phase Transitions), Feedback

Core Idea¶

Hysteresis is the property of a system whereby its current state depends not only on the current external conditions but also on the path by which those conditions were reached ^[1], so that the system's response curve to a changing parameter forms a loop rather than a single-valued function. The essential commitment is that history is encoded in the system's internal state in a way that the external parameter alone cannot reveal: returning the parameter to a prior value does not return the system to its prior state. Every hysteresis claim specifies (1) the system and the state variable being tracked, (2) the external parameter being varied, (3) the path-dependence — different state values at the same parameter value depending on history — and (4) the internal mechanism (multiple stable states, internal adjustment lag, irreversible structural change) that encodes the history.

How would you explain it like I'm…

The Paperclip Remembers

Bend a paperclip a little and let go; it springs back. Bend it a lot and let go; it stays bent. Where the paperclip ends up depends on what you did to it before, not just where your hand is now. The paperclip remembers.

When a System Remembers Its Past

Hysteresis is when a system's state depends on its history, not just on what's happening right now. Think of your thermostat: it turns the heater on when the room drops to 68 and off at 72, so at exactly 70 degrees it could be either on or off depending on whether the room was warming up or cooling down. The system 'remembers' the path it took to get to this point, so going back to the same temperature doesn't always undo the change.

Path-Dependent System State

Hysteresis is the property where a system's current state depends not only on the present external conditions but also on the path by which those conditions were reached. If you graph the system's response while raising and then lowering some input, you don't trace a single line; you trace a loop. A classic example is magnetizing iron: increase the magnetic field and the iron magnetizes; remove the field and it stays partly magnetized rather than returning to zero. The system has internal memory, sometimes from multiple stable states, sometimes from internal lag, sometimes from irreversible structural change.

Hysteresis is a property of certain dynamical systems whereby the present state depends on the history of the inputs, not merely their current value, so the response curve to a cyclically varied parameter forms a loop rather than a single-valued function. The system carries internal state (hidden from the external parameter) that encodes information about the trajectory taken. Classic examples include ferromagnetic hysteresis (iron's magnetization lags the applied field, tracing a B-H loop), elastic hysteresis (rubber's stress-strain curve differs on loading versus unloading), and economic hysteresis (unemployment persistently elevated after a recession ends, because worker skills and firm networks degrade irreversibly). Every hysteresis claim must specify the state variable being tracked, the external parameter being varied, the observed path-dependence (different states at the same parameter value), and the internal mechanism, typically one of: multiple stable equilibria, adjustment lag, or irreversible structural change.

Structural Signature¶

A relationship exhibits hysteresis when each of the following holds:

State variable. A measurable property of the system — magnetization, employment rate, ice cover, ecosystem composition — is tracked.
External parameter. A control parameter (applied magnetic field, GDP growth rate, temperature, harvesting pressure) is varied and can be increased and decreased.
Path-dependent response. As the parameter is varied through a cycle, the state variable traces different values during increase and decrease — the response curve forms a closed loop. The response depends on the history of parameter variations, not just current values ^[2].
Multiplicity at single parameter value. At intermediate parameter values, the state can take more than one value depending on the history of the parameter trajectory ^[3].
Internal memory mechanism. Some internal feature — multiple stable equilibria, irreversible structural changes, time-lagged internal states — preserves history beyond what the external parameter encodes.
Loop area as energy or cost. The area of the hysteresis loop typically corresponds to energy dissipated or cost imposed by cycling the parameter, distinguishing reversible from dissipative dynamics ^[4].

What It Is Not¶

Not simple lag. A first-order linear lag produces a phase shift in the response but retains a single-valued steady-state relationship; hysteresis produces multiple steady states at the same parameter value. Lag and hysteresis often coexist but are distinct.
Not irreversibility per se. Irreversibility concerns whether a process can be undone in principle; hysteresis concerns whether the state returns when the parameter returns. Some hysteretic loops are nondissipative (idealized magnetic hysteresis at zero temperature has limits); some irreversible processes do not show hysteresis. The two overlap but differ. See irreversibility.
Not a tipping point. A tipping point is a critical parameter value where the system jumps between regimes; hysteresis is the phenomenon that the jump-up and jump-down values differ, producing a loop with two thresholds. Hysteresis is one signature of a multi-stable system with tipping points; the prime concepts are linked but distinct. See tipping_points_or_phase_transitions.
Not memory in the cognitive sense. A hysteretic system encodes history in its physical state without representing or retrieving it; cognitive memory involves representational content. The structural pattern of "history in current state" is shared, but the mechanisms differ.
Not chaos. Chaotic systems exhibit sensitive dependence on initial conditions but remain on a single attractor; hysteretic systems have multiple attractors selected by history. They are different kinds of history-sensitivity.
Common misclassification. Calling any delayed response "hysteresis"; treating any irreversibility as hysteresis; asserting hysteresis without evidence of two-valued state at the same parameter value.

Broad Use¶

Physics and materials science
- Magnetic hysteresis in ferromagnets (B-H curves) ^[5]; ferroelectric hysteresis; mechanical hysteresis in elastic materials; superconductor flux trapping ^[5].
- Ferromagnetic domain switching and coercive-field determination ^[6]; Steinmetz exponent for power-loss engineering ^[6].
Engineering and control
- Backlash in mechanical linkages; hysteretic damping; thermostat hysteresis preventing rapid cycling; relay hysteresis ^[7].
- Bouc-Wen models for nonlinear vibration ^[8]; shape-memory alloy design ^[9].
Climate and Earth science
- Ice-sheet hysteresis under temperature forcing; vegetation-climate coupling with multiple stable states; glacier mass balance.
Ecology
- Ecosystem hysteresis in lake eutrophication ^[10]; coral reef regime shifts; rangeland degradation; population dynamics with Allee effects.
Economics and labor markets
- Hysteresis in unemployment after recession ^[10]; persistence of trade patterns; monetary regime hysteresis; inflation expectations ^[10].
Biology and physiology
- Cell-fate hysteresis in development; hysteresis in neural firing thresholds; membrane potential hysteresis in pacemaker cells.
Nanotribology and friction
- Molecular-scale hysteresis as friction mechanism ^[11]; adhesion hysteresis; stick-slip dynamics.

Clarity¶

Hysteresis clarifies by separating two questions that vague "lag" or "stickiness" language conflates: does the steady-state relationship between parameter and state actually depend on direction of approach, and what internal mechanism produces that path-dependence? A claim like "unemployment is sticky after recessions" resolves into "the steady-state unemployment rate as a function of macroeconomic conditions exhibits hysteresis: at the same conditions, an economy approaching from a recession has higher unemployment than one approaching from an expansion; the internal mechanism involves skill atrophy, network erosion, and changed employer-side filtering." The clarifying force is to distinguish reversible-but-slow from irreversible-without-additional-action, and to identify what specifically encodes the history ^[12].

Manages Complexity¶

Replaces single-valued response models with loop dynamics that capture realistic multi-state behavior.
Supports prediction of asymmetric response: identifying the upward and downward thresholds tells the analyst when a system will switch regimes and when it will resist returning.
Guides intervention: knowing the hysteresis loop allows targeting actions to threshold approaches (preventing transitions) or to mechanism removal (escaping a stuck regime).
Enables design: hysteretic switching is used deliberately in thermostats, Schmitt triggers, and shape-memory alloys to provide robust state holding.
Connects to multiple-equilibria analysis: hysteresis is the dynamical signature of bistability or multi-stability and informs bifurcation-theoretic models.

Abstract Reasoning¶

Hysteresis trains a reasoner to ask:

Does the system's state depend on the path by which the parameter reached its current value?
What internal mechanism encodes the history — multiple stable equilibria, structural irreversibility, internal lag, network reconfiguration?
What are the upward and downward thresholds of the hysteresis loop, and how wide is the loop?
What energy or cost does the loop area represent?
Can the loop be narrowed or eliminated by changing system structure (adding noise, reducing a positive-feedback loop, removing the bistability mechanism)?
Is the hysteresis a feature (robust state holding, dead-band stability) or a bug (stuck-in-bad-state, irreversible damage)?

Knowledge Transfer¶

Role mappings across domains:

State variable ↔ magnetization / unemployment rate / ice cover / lake trophic state / vegetation cover / cell fate
External parameter ↔ applied field / output gap / temperature / nutrient loading / herbivore pressure / signaling concentration
Loop width ↔ coercive field / threshold asymmetry / lag distance
Internal memory mechanism ↔ domain structure / skill loss / albedo feedback / internal nutrient pool / network reconfiguration / gene-regulation switch
Upward threshold ↔ flipping point / collapse threshold / activation level
Downward threshold ↔ recovery point / reset level / restoration threshold
Loop area ↔ dissipated energy / switching cost / persistent welfare loss
Multistability ↔ alternative stable states / regime alternatives / discrete cell fates

A materials scientist measuring magnetic hysteresis loops, an ecologist tracking lake eutrophication and recovery, and a labor economist analyzing unemployment persistence are all doing the same structural work: identify the state and parameter, document the path-dependent response, characterize loop width and energy, and identify the internal memory mechanism. The same diagnostic — "what state, what parameter, what loop, what mechanism encodes history?" — applies across their contexts, with the same failure modes (treating loops as lag, missing multistability, mistaking the mechanism) in each.

Examples¶

Formal Example: Ferromagnetic Hysteresis¶

Physics. Magnetic hysteresis in a ferromagnetic material. State variable: magnetization M. Parameter: applied magnetic field H. Path-dependent response: cycling H from positive to negative and back traces a closed loop in M-H space, not a single curve. Multiplicity: at H = 0 the material has remanent magnetization +Mr or −Mr depending on prior history. Internal mechanism: domain structure with energy barriers between configurations; coercive field Hc must be exceeded to flip domains ^[13]. Loop area: corresponds to energy dissipated per cycle (heat). Every item of the structural signature is operative and well- characterized quantitatively.

Ferromagnetic hysteresis arises from the pinning of magnetic domain walls at structural defects and interfaces. The Preisach model ^[14] provides a mathematical framework decomposing the macroscopic loop as a superposition of elementary hysteron units, each exhibiting two-state transitions at thresholds. This formalism explains minor loops and irreversible magnetization changes.

Mapped back: Ferromagnetic B-H curves exemplify the complete hysteresis structural signature: state (M), parameter (H), two-valued remanence, energy loss per cycle, and a well-characterized mechanism (domain reversal at threshold fields).

Applied Example: Economic Hysteresis in Labor Markets¶

Non-physical, structurally faithful. Hysteresis in unemployment persistence after recession. State variable: unemployment rate. Parameter: output gap (deviation from potential GDP). Path- dependent response: unemployment rises sharply when GDP falls below potential (recession); but falls more slowly when GDP recovers, and often fails to return to pre-recession levels even when output regains its potential. Multiplicity: at the same output gap, the economy approaching from recession has higher unemployment than one entering from expansion. Internal mechanism: skill atrophy, erosion of job-search networks, employer-side screening changes, and depreciation of human capital during unemployment ^[10]. Loop area: sustained welfare loss and foregone output during recovery phase.

The hysteresis arises from irreversible changes in the labor force: workers displaced by recession accumulate skill obsolescence, break professional networks, and face higher hiring barriers on re-entry. These feedback mechanisms—positive feedback sustaining unemployment—prevent simple reversal when aggregate demand recovers.

Mapped back: Economic unemployment hysteresis embodies the same structural commitment as ferromagnetic hysteresis: dual thresholds (recession trigger vs recovery threshold differ), energy/welfare cost in loop area, and a mechanism (human-capital loss) encoding history in steady-state outcomes.

Structural Tensions and Failure Modes¶

T1 — Path-Dependent vs Path-Independent State:

Structural tension: Hysteresis is fundamentally path-dependent; the state at a given parameter value varies with history. However, dynamical systems at equilibrium are path-independent: the final state depends only on parameter values, not on the trajectory taken. Some systems exhibit bistability and hysteretic loops at fast cycling rates but converge to single-valued steady states at slow cycling (relaxation to equilibrium). Distinguishing genuine path-dependence (encoded in stable structure) from slow approach to a path-independent steady state requires careful design of experiments and understanding of timescales.
Common failure mode: Declaring hysteresis from fast cycles without verifying the loop persists at equilibration timescales; conversely, dismissing apparent path-dependence as mere lag when structural multistability actually encodes history.

T2 — Static vs Dynamic Hysteresis (Rate-Independent vs Rate-Dependent Loops):

Structural tension: Classical hysteresis models (Preisach, Ewing, Maxwell) assume rate-independence: the loop depends only on the range traversed, not the speed of cycling. However, real systems exhibit rate-dependent loops: faster cycling produces wider loops (more dissipation per cycle at higher rates) or narrower loops (if viscous relaxation cannot keep pace). Ferroelectric ceramics, polymeric materials, and biological systems often show rate-dependent behavior. Should hysteresis theory prioritize idealized rate-independent models or empirically accurate rate-dependent treatments?
Common failure mode: Applying rate-independent models to rate-dependent systems without calibrating for speed; misinterpreting loop-width changes across cycling frequencies as unphysical artifacts rather than real rate-dependence.

T3 — Major Loops vs Minor Loops (Preisach Decomposition Assumptions):

Structural tension: The Preisach model constructs hysteresis as a superposition of elementary two-state hysteron units, each triggered by threshold-crossing. Experimental major loops (full-range cycling) are well-described this way. However, when the parameter undergoes reversals without reaching extreme values (minor loops), Preisach predictions can diverge from observation. Real materials exhibit "memory" of the full-range extrema (wiping-out effect), which pure Preisach does not capture. Should hysteresis theory prioritize mathematical elegance (Preisach) or empirical accuracy of minor-loop behavior?
Common failure mode: Using Preisach decomposition without verifying it against measured minor loops; missing historical-maximum-dependent effects.

T4 — Microscopic Mechanism vs Macroscopic Phenomenology (Multiple Substrates):

Structural tension: Hysteretic loops in different materials arise from distinct mechanisms: ferromagnetism (domain reversal), elastic materials (dislocation motion), polymers (molecular-chain reorientation), superconductors (flux quantum pinning). Yet all produce qualitatively identical macroscopic B-H or stress-strain signatures. Is the right level of explanation the mechanism (specific to each system) or the universal hysteresis form? Can a unified theory capture all mechanisms, or must each be studied separately?
Common failure mode: Over-generalizing from one mechanism (e.g., magnetic domains) to predict behavior in a different substrate (e.g., friction); missing substrate-specific transitions that break hysteretic linearity.

T5 — Beneficial vs Harmful Hysteresis (Memory as Feature or Bug):

Structural tension: Hysteresis is deliberately engineered into Schmitt triggers, thermostats, and shape-memory alloys to provide noise-immune switching and state-holding. In these applications, hysteresis is a design feature. Conversely, hysteresis in economic systems (unemployment traps), ecosystems (eutrophication), and climate (ice-sheet bifurcations) represents persistence of undesired states. The same structural property—history encoding—is beneficial in engineered systems but harmful in natural ones. Removing hysteresis from control systems can introduce instability; enforcing hysteresis in labor markets reduces flexibility. How do we distinguish desirable from undesirable hysteresis, and can we alter it selectively?
Common failure mode: Reflexively treating hysteresis as undesirable in all contexts and attempting to eliminate it; or conversely, engineering robust hysteresis without anticipating downstream rigidity costs.

T6 — Classical vs Quantum Hysteresis (Decoherence, Many-Body Localization, Spin Glasses):

Structural tension: Classical hysteresis assumes distinguishable, semiclassical trajectories in phase space. Quantum mechanics introduces fundamental indistinguishability and superposition: at the quantum scale, a system can occupy multiple states simultaneously. At what scale does hysteresis emerge from quantum mechanics? Spin-glass systems exhibit history-dependence and complex energy landscapes, but the role of quantum tunneling, decoherence, and many-body localization in producing or suppressing hysteresis is subtle. Do quantum systems exhibit "true" hysteresis, or does decoherence create the appearance of classical hysteresis?
Common failure mode: Applying classical hysteresis concepts to quantum systems without accounting for coherence and tunneling; misinterpreting quantum metastability (long coherence times without dissipation) as hysteresis with memory.

Structural–Framed Character¶

Hysteresis sits at the structural end of the structural–framed spectrum: it is a pure relational pattern, the same in any domain where it appears, and nothing about its meaning depends on a particular field's vocabulary or assumptions.

The definition is wholly formal: a system whose current state depends on the path by which conditions were reached, so its response curve forms a loop rather than a single-valued function. No home vocabulary needs to travel, because there is none to import — the same pattern describes magnetization in a metal, employment rates in an economy, ice cover in a climate, and species composition in an ecosystem without changing meaning. It carries no built-in evaluative weight, owes nothing to human institutions, and can be specified entirely in terms of state variables and external parameters. Encountering it is recognizing a structure already present in the system, not importing a perspective onto it. On every diagnostic, it reads structural.

Substrate Independence¶

Hysteresis is a highly substrate-independent prime — composite 4 / 5 on the substrate-independence scale. Its structure — a system's current state depending on its path as well as present conditions, producing characteristic loop behavior with lag — is stated in medium-neutral terms of state variables, external parameters, and path-dependence. It recurs genuinely in magnetism, in labor markets where unemployment persists, in ecosystems with regime shifts and alternative stable states, and in material fatigue. Both breadth and abstraction sit at 4; what holds it short of universal is that the brief's examples likely stay within physics and economics, so the demonstrated transfer trails the principle's evident cross-substrate scope.

Composite substrate independence — 4 / 5
Domain breadth — 4 / 5
Structural abstraction — 4 / 5
Transfer evidence — 3 / 5

Relationships to Other Primes¶

Parents (2) — more general patterns this builds on

Hysteresis is a kind of Path Dependence

Hysteresis is a specialization of path dependence. The general path-dependence pattern says current state depends on the historical trajectory, not just on current external conditions. Hysteresis specializes by adding a particular signature: as an external parameter is varied up and down, the system's state traces a loop rather than a single-valued curve, with multiple stable states accessible at the same parameter value depending on history. The same history-encodes-state commitment persists, with the loop in the response curve as the diagnostic shape of the path dependence.
Hysteresis is a kind of State and State Transition

Hysteresis is a specialization of state and state transition. The general pattern specifies a state space and a transition relation, with future behaviour depending on current state plus inputs. Hysteresis instantiates this with multiple stable states existing at the same external parameter value: which one the system occupies depends on the path through parameter space, so the response curve forms a loop rather than a single-valued function. History is encoded in the system's internal state in a way external parameters alone cannot reveal, which is exactly the state-as-Markovian-summary structure the parent pattern names.

Path to root: Hysteresis → State and State Transition

Neighborhood in Abstraction Space¶

Hysteresis sits in a sparse region of abstraction space (73^rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely rather than landing on a neighbor.

Family — Dynamical Regimes & Tipping Points (11 primes)

Nearest neighbors

Coupling — 0.79
Tipping Points (or Phase Transitions) — 0.78
Inertia — 0.77
Regime Change — 0.76
Conservation Laws — 0.76

Computed from structural-signature embeddings · 2026-05-29

Not to Be Confused With¶

Hysteresis must be distinguished from Inertia, its closest structural neighbor. Both concepts involve systems that resist rapid change and encode history in their current state, but they operate on different timescales and mechanisms. Inertia describes resistance to changes in trajectory or motion—a mass in motion persists in that motion unless acted upon by a force; a system in a particular configuration resists acceleration or deceleration independent of how it arrived at that configuration. Critically, a system with inertia will eventually respond to sustained external forces and return to its prior state once the force is removed, because inertia itself does not create multiple stable states. Hysteresis, by contrast, is specifically about path-dependent steady states—the same external parameter value produces different outcomes depending on the history by which that value was reached. An undamped pendulum exhibits inertia (it swings past equilibrium because momentum carries it); it does not exhibit hysteresis (returning to the original angle restores the original state). A ferromagnet exhibits hysteresis (at zero applied field, magnetization depends on prior history); it exhibits inertia only at short timescales during acceleration. A system can have pure inertia without hysteresis; hysteresis always involves some internal mechanism (bistability, friction, structural change) that creates multiple stable states. The distinction clarifies why simple time delays or sluggish response (consequences of inertia) differ from locked-in path-dependence (the signature of hysteresis).

Nor is hysteresis equivalent to Equilibrium or Multiple Equilibria. Equilibrium describes a state where net forces balance and no spontaneous change occurs; equilibrium can be stable (the system returns to equilibrium if perturbed), unstable (small perturbations grow), or neutral (perturbations don't decay). Multiple equilibria exist when a system has several distinct states in which forces balance. However, multiple equilibria alone do not imply hysteresis: if the equilibria are path-independent—the system settles into a particular equilibrium determined solely by the current external parameter value, regardless of how that value was approached—then no hysteresis occurs. For example, a ball rolling into one of three potential wells on a landscape (different equilibria for different parameter values) exhibits multiple equilibria but not hysteresis if each parameter value uniquely determines which well the ball inhabits. Hysteresis requires that at intermediate parameter values, multiple equilibria coexist and which one the system occupies depends on history. This distinction matters for prediction: knowing a system's equilibrium doesn't reveal its actual state if hysteresis is present; knowing its parameter value and equilibrium type are insufficient without also knowing the history.

Hysteresis is also fundamentally distinct from Resonance, which concerns frequency-dependent amplification of inputs. Resonance asks: at what driving frequencies does the system respond most strongly? The answer depends on the system's stiffness and damping, but not on the path by which the driving frequency was reached. A wine glass resonates at a particular frequency regardless of whether that frequency arrived through gradual increase or sudden jump. Hysteresis, by contrast, asks: at the same parameter value, what state does the system occupy given its history? Resonance is frequency-domain; hysteresis is history-domain. Both involve amplitude-response curves, but resonance curves are typically single-valued functions of frequency, while hysteretic loops are multi-valued functions of parameter with different upper and lower branches depending on approach direction. A system can amplify resonantly without exhibiting hysteresis; a hysteretic system does not necessarily have pronounced frequency-dependent resonance. Confusing these concepts leads to misdiagnosis: treating hysteretic damping (energy dissipation in the loop) as resonance-peak enhancement, or vice versa.

Solution Archetypes¶

Solution archetypes in the catalog that build on this prime — directly (this prime is a source ingredient) or as a related prime.

Built directly on this prime (2)

Also a related prime in 15 archetypes

▸ Show 5 more

References¶

[1] Maxwell, James Clerk. A Treatise on Electricity and Magnetism, Vol. 2. London: Clarendon Press, 1873. Early systematic observations of magnetization-field relationship in ferromagnets; documents hysteretic behavior in magnetization curves; precursor to quantitative hysteresis models. ↩

[2] Ewing, James Alfred. "On the Production of Transient Currents in Iron and Steel Conductors by Twisting." Proceedings of the Royal Society of London, vol. 33 (1882): 21–23. Ewing coins the term "hysteresis" (from Greek "hysterein" — to lag behind) for the magnetic lag phenomenon in ferromagnetic systems; establishes foundational terminology and recognizes it as universal physical process. ↩

[3] Ewing, James Alfred. "Experimental Researches in Magnetism." Philosophical Transactions of the Royal Society, vol. 176 (1885): 523–640. Comprehensive experimental characterization of ferromagnetic hysteresis loops; systematic variation of field amplitude and frequency; establishes empirical foundation for hysteresis quantification. ↩

[4] Steinmetz, Charles Proteus. "On the Law of Hysteresis." Transactions of the American Institute of Electrical Engineers, vol. 9 (1892): 3–64. Quantitative engineering model for magnetic hysteresis energy loss; derives B^1.6 power law for core loss; establishes engineering framework for AC power system design and transformer efficiency calculation. ↩

[5] Preisach, Franz. "Über die magnetische Nachwirkung." Zeitschrift für Physik, vol. 94 (1935): 277–302. Develops Preisach model: hysteresis as superposition of elementary two-state hysteron units; provides rigorous mathematical framework for decomposing complex hysteresis loops; foundation for modern hysteresis theory. ↩

[6] Bertotti, Giorgio. Hysteresis in Magnetism: For Physicists, Materials Scientists, and Engineers. San Diego: Academic Press, 1998. Comprehensive treatment of magnetic hysteresis with statistical mechanics framework; covers domain structure, energy dissipation, and phenomenological models; bridges microscopic and macroscopic scales. ↩

[7] Wen, Yi-Kwei. "Method for Random Vibration of Hysteretic Systems." Journal of the Engineering Mechanics Division (ASCE), vol. 102, no. 2 (1976): 249–263. Extends Bouc model (Bouc-Wen formulation); provides engineering standard for seismic analysis of structures with hysteretic damping; widely adopted in civil-engineering design codes. ↩

[8] Bouc, Robert. "Modèle mathématique d'hystérésis." Acustica, vol. 24 (1971): 16–25. Develops Bouc model: differential-equation formulation of hysteresis for nonlinear vibration; enables integration with dynamical systems theory; engineering-oriented approach. ↩

[9] Mayergoyz, Isaak Davydovich. Mathematical Models of Hysteresis. Berlin: Springer, 1991. Definitive monograph on mathematical hysteresis; includes Preisach, vector models, and rigorous functional-analysis treatment; standard reference for theoretical hysteresis. ↩

[10] Heine, Bastian, and Klaus Müller. "Hysteresis in Economic Relationships: An Overview." Empirical Economics, vol. 22 (1997): 309–345. Systematic review of hysteresis in labor markets, inflation, trade patterns, and macroeconomics; documents empirical prevalence; contrasts with neoclassical path-independence assumptions. ↩

[11] Tomlinson, Geoffrey William. "A Molecular Theory of Friction." Philosophical Magazine, vol. 7, no. 48 (1929): 905–939. Proposes a microscopic picture of friction as the dragging of atoms and molecules over potential-energy barriers; early attempt to connect macroscopic friction to molecular-scale mechanisms; precursor to modern tribology and microscopic damping theory. ↩

[12] Visintin, Augusto. Differential Models of Hysteresis. Berlin: Springer, 1994. Modern PDE treatment of hysteresis as nonlocal memory operator; extends classical phenomenology to partial differential equations; bridges physics and mathematics. ↩

[13] Wischert, Gerrit, Alexander Wunderlin, Bernd Pelster, Andreas Olivares, and Adi Bulsara. "Indistinguishability of Stochastic and Deterministic Dynamics in Escape over a Fluctuating Barrier." Physical Review E, vol. 49, no. 5 (1994): R4365–R4368. Demonstrates equivalence between stochastic noise-driven transitions and deterministic chaos in barrier-crossing dynamics; clarifies distinction between stochastic and deterministic hysteresis; shows when noise masks or enhances hysteretic behavior. ↩

[14] Mayergoyz, Isaak Davydovich. "Mathematical Models of Hysteresis." Physical Review Letters, vol. 56, no. 15 (1986): 1518–1521. Announces comprehensive generalization of Preisach model; extends framework to multi-dimensional parameter spaces and nonlinear materials. ↩