Claude Elwood Shannon

Record year1948

M rationaleM7 is adopted: the original measures of information, channel capacity, and limits of reliable transmission have been used across generations in teaching and to guide communication/storage problems, consistent with the continuing contribution of a scientific foundation on the same basis as Maxwell’s original field equations. M6 would recognize only the disciplinary foundation delivered by the first paper and would not sufficiently represent the available evidence of continued use; M8 requires a stronger boundary of civilization-level results and is not taken here. Subsequent new codes, devices, and modern user utility are accounted for separately.

attribution rationaleResearch point estimate: the original paper explicitly takes Nyquist/Hartley as a foundation of communication theory and uses logarithmic information, while also explicitly extending noise, the statistical structure of messages, and the properties of the destination, and completing a unified construction of entropy, capacity, and reliable communication in the presence of noise. Nyquist/Hartley 0.14 is retained as an identifiable predecessor attribution; the generalized mathematics pool is narrowed to inputs actually used, such as stochastic processes, Markov processes, and ergodic theory, at 0.05; the remainder with no identified contributing actions is reduced to 0.05; the released share is attributed to Shannon’s work for unification, generalization, and proof, with a central value of 0.76.

overlap boundaryM counts only the sustained knowledge effect, through the observation date, of the original measures of information, capacity, and limits of reliable transmission; subsequent coding, devices, industrialization, and new instructional designs are assessed separately. The 1959 rate-distortion work and similar results count only objects newly added beyond the 1948 framework.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works. Direct cross-generational-use evidence for high M grades must show source locations just as philosophical contributions must; familiar conclusions from the history of science do not exempt it from verification. Re-crediting named predecessors is likewise subject to a review for previously credited contributions. Re-estimation of attribution in this round: precise proportions are not mechanical conversions of citation counts or paper authorship. The full original works of Nyquist, Hartley, Wiener, and the relevant mathematical prehistory have not yet been compared item by item, nor has it been shown whether colleagues' suggestions altered the core theorems; therefore sole credit is not assigned to Shannon. If predecessors already contain the key definitions, theorems, or proof structures of this object, the corresponding credit should be removed from Shannon's share; general mathematical tools, institutional setting, publication, teaching, and dissemination do not receive formation shares.

Record year1938

M rationaleThe M6 disciplinary-foundation anchor is adopted: the two-way correspondence between logical expressions and relay networks, computable simplification, and synthesis from specifications to circuits form a complete design language that directly changes the capacity to express and solve this specialized problem. M5's single international method does not adequately describe this foundational structure; this assessment does not claim subsequent computers, chips, or all industrial utility, nor does it erase the delivered system because the most general minimization problem remains unsolved.

attribution rationaleShannon established a two-way correspondence between logical expressions and relay networks and supplied procedures for analysis, simplification, and synthesis, so the principal formation action is assigned 0.75; symbolic logic is a specific structural input at 0.15, network theory at 0.05, and the remaining 0.05 is unallocated.

overlap boundaryCount only the 1938 logic-relay two-way mapping, analysis, and initial synthesis; CAT050's subsequent extended synthesis constructions and bounds, machine-assisted processes, and the computer industry are accounted for separately.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1959

M rationaleM5.3: beyond the known 1948 idea of fidelity, R(d) and the asymptotic achievability/converse theorems unify source coding with permitted distortion into a computable, testable rate-fidelity boundary, reaching a foundational reconstruction of an international field method. The depth of its unified problem and solution places the central estimate above the general important-method grade M5; its limited conditions and the pre-existing 1948 framework mean that this increment does not take credit for all communication/compression capacity or the entire scientific foundation. This judgment requires no additional evidence of deployment and does not add back actual codecs or later extensions.

attribution rationaleShannon introduced R(d) and delivered the principal new content of the achievability and converse theorems, assigned 0.82; the 1948 framework is an already-consumed predecessor and cannot be used to allocate Shannon another share of credit in this item, while other formation inputs not yet separated remain at 0.18.

overlap boundaryCount only the increment of R(d), the fidelity criterion, and the corresponding direct and converse theorems relative to the 1948 framework; do not count modern codecs, the compression industry, or the 1948 general theory of capacity.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

summaryA published mathematical framework for discrete finite-state sources: information elements as reversible finite-state coding equivalence classes; conditional-entropy distance, inclusion, joint and common information operations, and explicit non-distributive examples.

M rationaleM5.2 is retained: equivalence classes, distance, inclusion, and joint/common-information operations for finite-state information constitute a complete international method, and chains of subsequent theoretical use in 2007 and 2023 support its sustained specialist effect; it remains below a disciplinary foundation and does not include later common-information theory.

attribution rationaleShannon's delivery of the framework, operations, and counterexamples is assigned 0.70; specific structural inputs from entropy theory and lattice theory receive 0.20, and unseparated formation inputs remain at 0.10.

overlap boundaryCount only the method of finite-state information lattices; the 1948 entropy and coding framework, and later common-information results, are not counted again.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1949

M rationaleM5 is adopted: the public structure of messages, ciphertexts, keys, and probabilities refines secrecy criteria, equivocation, and key conditions, delivering a complete and important international theoretical method. The scope reviewed so far chiefly establishes the content and origin of this conditional knowledge; it is still insufficient to claim, under M6, its own broader sustained effect as a disciplinary foundation. Modern security systems are not backfilled, and the knowledge qualification is not reduced for lack of deployment.

attribution rationaleShannon's completion of the public secrecy structure, criteria, and derivations is assigned 0.72; existing cryptographic systems and probabilistic tools, as direct material for this new structure, receive 0.20, and the remaining 0.08 is unallocated.

overlap boundaryCount only the public mathematics of secrecy from 1949; the 1948 general measures of information and capacity are not counted again, and modern security systems and deployments are not backfilled.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleStrict zero-error results and adjacency representations, evaluable bounds, and feedback evaluation constitute an operationally usable, important international method of knowledge. An independent study in 1978 describes its unsolved instances; the 2026 research in subsequent materials lies beyond this task's effect cutoff and serves only as a lead for later verification; proofs by new authors and actual deployment are not counted. A grade below 5, namely 4, may be an acceptable alternative judgment with a narrower scope; grade 6 requires a broader scope of global capability/disciplinary foundation. This object bears only the method of the zero-error subfield and does not include information theory as a whole from 1948. It is not reverse-engineered from Shannon's prior score or a target score.

attribution rationaleShannon's completion of the principal zero-error object, adjacency formulation, bounds, and examples is assigned 0.68; Elias's key observation about feedback gains and suggestions for the proof receive 0.22; the remaining 0.10 is retained.

overlap boundaryCount only the new method of zero-error capacity, adjacency representations, and feedback evaluation; the 1948 general theory of capacity, subsequent graph-theoretic results, and deployment are not counted again.

Evidence gaps The effect cutoff for this assessment is 2025-12-31; 2026 use instances are not adopted, and M5 remains supported by the content of the original knowledge and the record of acceptance before the cutoff.

Record year1956

M rationaleTwo independent simulation constructions jointly deliver the model invariance of state/symbol tradeoffs; they more directly alter the general description of computational models than a local sharp bound does, but do not claim credit for the computer industry.

attribution rationaleShannon's delivery of two classes of simulation constructions for state/symbol tradeoffs is assigned 0.72; Turing's universal-machine object is an irreplaceable specific predecessor input at 0.20, and the remaining 0.08 is unallocated.

overlap boundaryCount only the simulation constructions under state/symbol constraints; Turing's original universal-machine result, subsequent minimal machines, and the computer industry are not counted again.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1941

M rationaleThis result elevates an existing machine into a general theoretical object whose capability limits can be discussed, and it has been continuously studied in subsequent technical literature, exceeding an ordinary result within a single specialist subfield. However, a gap was later identified in its key proof, and the demonstrated effect is concentrated in analog-computation theory; this is still insufficient to assign M5 directly under the anchor for a complete international platform/field change. M4.5 is an explicit normative point between M4 and M5 and is not elevated by the author's reputation, later corrective proofs, or hardware benefits.

attribution rationaleShannon's elevation of the machine into a mathematical object of capability limits and completion of the main analysis is assigned 0.65; the specific mechanical and problem inputs from Bush/MIT's existing analyzers receive 0.20, Guerrieri's function generation 0.07, Kelvin's prior conception 0.03, and the remaining 0.05 is unallocated.

overlap boundaryInclude the unified mathematical modeling of the idealized differential analyzer; solvability/generability problems under the paper's assumptions; continuous-function approximation; and gear-ratio and speed-control analysis. Do not count the invention of the physical analyzer, later computing devices and applications, or portray a main theorem with a gap, or later corrective proofs, as results already completed by Shannon.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1949

summaryPublished long-form treatment frames two-terminal switching-circuit synthesis as finding circuits with given behaviour/best realization and develops general methods, constructions and bounds beyond merely introducing Boolean analysis.

M rationaleUnder the condition of independent isolation, the extended work defines two-terminal switching-circuit synthesis as problems of behavioral realization and optimal realization, and supplies general methods, constructions, and bounds; it is a complete field method. The 4.4 assessment is based on the structural capacity of a complete problem framework and multiple classes of methods; it is above a single theorem, but its scope remains a specialist subfield and overlap review is incomplete, so the central estimate remains in the M4 grade. The current summary does not identify the distinct outcome separating this work from earlier Boolean methods and machine-assisted design, and the object is limited to two-terminal switching circuits; it cannot enter M5 or above merely by the label “synthesis.”

attribution rationaleShannon's completion of general synthesis constructions and bounds for two-terminal switching circuits is assigned 0.75; Boolean input receives 0.08, other logical/switching prehistory 0.12, and the remaining 0.05 is unallocated. Publication in BSTJ is not a method-forming action.

overlap boundaryThe 1938 item already counts the logic-relay two-way mapping; this item counts only the subsequent synthesis constructions and bounds, while machine-assisted processes are listed separately.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleCentral M4.2: the proof that complete feedback still cannot increase the capacity of an ordinary DMC rules out a path to capacity gain across an entire class of communication models and separates capacity from gains in reliability/complexity; this is a durable specialist design and theoretical boundary, in the lower part of the M4 grade, and is not expanded into a cross-model foundation of communication theory.

attribution rationaleShannon's proposal and proof that noiseless feedback does not increase capacity for ordinary DMCs is assigned 0.85 for the principal formation responsibility; inputs and alternative paths that have not been linked to specific contributing actions remain at 0.15.

overlap boundaryCount only the boundary that noiseless feedback does not increase capacity for ordinary DMCs; the 1948 general capacity, feedback's reliability/complexity benefits, and other channel models are not counted again.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1957

M rationaleCentral M4.1: it gives usable coding bounds among finite block length, input distribution, and error, answering performance constraints not provided by the 1948 asymptotic existence conclusion; the central tool still has a clear place in derivational comparisons with stronger modern bounds. Its added scope is the family of finite-DMC methods rather than all non-asymptotic theory, and it is placed at the starting range of M4.

attribution rationaleShannon constructs random coding, average-error bounds, and decoding relations; Chernoff exponential tail bounds support only local analysis, while unrecorded prior contributions and the timing of alternative development remain unresolved. The central 88%/5%/7% is a normative accounting allocation.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1959

M rationaleThe entire equal-energy Gaussian-channel family of error bounds (conic-geometric lower bounds, random spherical-code upper bounds, matching above the critical rate, and finite-length estimates) is a single outcome; M=4.2. It is not split into multiple items, and neither low rates, arbitrary power, continuous time, nor industrial benefits are counted.

attribution rationaleShannon 78%, coding geometry and statistical prehistory 12%, and unseparated support 10%; alternative budgets show the sensitivity of this normative allocation, without fabricating individual causal shares in the prehistory.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1961

M rationaleM4.1: treating interactions that one-way models cannot express through operationally meaningful two-way rate regions, inner and outer bounds, and code constructions, it is a complete set of theoretical tools for a line of research. Its value exceeds a single theorem example; it still does not encompass later network information theory as a whole, nor does it claim that a general exact single-letter capacity has been solved. The original candidate phrase “field change” here refers only to the two-way-channel research branch, rather than directly applying M5 for change across an entire international field.

attribution rationaleShannon completes the principal work; Blackwell contributes multiplicative examples and differences between bounds, Chernoff tail bounds support only parameter-error analysis, and the remainder is unknown.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1957

M rationaleM4.0: on the complete class of models with reconstructible internal states, it gives eigenvalue capacity, achievability, and converse bounds, expanding the objects that existing channel methods can handle rigorously and reaching the M4 starting point of relatively broad technical-method capability; its conditions of application remain strict, and it is not equated with general finite-state capacity or credited again with the 1948 foundational theory.

attribution rationaleShannon defines a model with states computable at both terminals, constructs eigenvalue capacity, and proves achievability and converse bounds; Frobenius's eigenvalue theory for nonnegative matrices is a specific tool input, while the remaining account retains unrecorded prehistory. The central 86%/8%/6% is a judgment of responsibility.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1948

summaryDevelops an algebraic network-ring treatment of graph topology: basis/normal-form theorems, factor/cut-set enumeration, duality and a construction procedure for Campbell’s network-realizability problem.

M rationaleProvided it qualifies as an independently counted outcome, network-ring algebra supplies bases, normal forms, factor/cut-set enumeration, duality, and a construction procedure for network realizability, forming a systematic specialist field method. The 4.2 assessment, based on theoretical systematicity and constructiveness, approaches the field-method grade; its domain remains a specialist subfield and deduplication awaits verification, so it is only slightly above the M4 anchor. The object is concentrated on graph topology, cut sets, and network realizability; its applicability to broader network science or engineering has not been established, and its boundaries with early Boolean analysis and edge coloring also await verification.

attribution rationaleShannon's completion of the new algebraic system receives 60%; Chinese research on determinants 15%; Foster's suggestion and Whitney's theory 10% each; and Bell 5%.

overlap boundaryCAT42 is network-ring algebra for graph topology; its technical object differs from both the 1938 Boolean-relay analysis and CAT44 edge coloring.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1955

summaryThe memorandum maps ideal rectifier circuits to posets, gives canonical normal form and rectifier-count bounds, presents duality for contact/rectifier networks, and describes a partially ordered-set analyzer.

M rationaleUnder the condition of separately counted outcomes, the outcome maps ideal rectifier circuits to posets, provides canonical forms, element-count bounds, contact/rectifier network duality, and an analyzer scheme, forming a systematic method. 4.0 is supported jointly by the theoretical system and construction/analysis capability, landing at the methodological anchor of the professional field; limits on applicability and risks of double-counting constrain a higher estimate. The analyzer is only part of the described construction; the applicability remains limited to ideal rectifier networks and poset models; its independence from earlier network algebra and switching synthesis has not been fully established.

attribution rationaleShannon takes 75% for the specific poset circuit theory; Birkhoff's lattice theory 10%, Moore's suggested construction 5%, Bell and circuit prehistory 5% each.

overlap boundaryCAT96's ideal rectifier poset, bounds, and duality are independent of CAT50 switching synthesis; the analyzer is counted only as described in the text, not as an already deployed implementation.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1958

M rationaleM3.8: The function-valued input-symbol approach transforms the problem of channel-state information causally available at the transmitter into an ordinary channel of equal capacity while preserving code error, providing a complete and reusable specialized method. It is deeper than a one-off local technique, yet remains a branch of specific state conditions in information theory; it falls in the upper segment of M3 professional sustained results, without borrowing unverified non-causal/continuous generalizations or industrial gains. The center is the responsibility judgment of this depth and coverage, not taking the midpoint because the interval is symmetric.

attribution rationaleShannon constructed function-valued input symbols and proved the equal-capacity transformation; absence would delay the core reduction. No evidence of named co-authors; the remainder is left unknown.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1942

M rationaleThe 73-page report forms a systematic analysis/synthesis framework for linear differential equation machines; the theorems and design procedures clearly exceed single local techniques, placing it at the high end of important sustained results in the professional subfield; but existing evidence establishes only the limited work delivered and its later inclusion in a published collection, not widespread adoption, systemic industry change, or fire-control effects, so it is below the M4 full anchor.

attribution rationaleShannon delivered the theory and design for linear differential equation machines; specific prehistory and unseparated engineering inputs remain in the budget; the share is normative attribution, not citation counting.

overlap boundaryOnly the newly delivered knowledge from the 1942 report is counted: linear machine non-degeneracy, orientation, sufficient gearing, reversibility, admittance analysis, duality, minimal design, and general design. Shared mechanical components and the 1941 general generation/solvability theory are causal inputs, not re-credited in this item's M; the entire fire-control system, specific equipment adoption, later GPAC theorems, or weapon operations are not counted.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1944

M rationaleProvides executable decision rules for the detection specialty, covering colored noise and unequal priors; narrower than general computational models and transforms, hence M=3.8. The numerical distance is a revisable value judgment.

attribution rationaleThe specific derivation and generalization were delivered by Shannon; Bayes' theorem and Gaussian quadratic forms are necessary statistical inputs. The center is a normative split of Shannon 72%, mathematical prehistory 23%, and unobserved communication/dissemination 5%, not by headcount or title.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1958

M rationaleM3.5: The channel inclusion relation with correlated pre- and post-randomization plus same-code-length error transfer links simulation, finite-code performance, and capacity into a reusable comparison tool, located in the mid-range of professional sustained results. It exceeds a single computation, yet has clear scope such as not comparing all channels; it does not count general lattice results/converse propositions or later applications as delivered.

attribution rationaleShannon defined randomized inclusion and proved error transfer; absence would delay unified comparison. Silverman is cited for context; this does not automatically establish a credit share.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1967

M rationaleM4.8: An admitted two-part combined DMC lower-bound method, including finite alphabets, list decoding, and conditional exponential structure, assigned a considered value estimate for its enduring contribution to converse results in the field. Scope excludes implementation codes, future improvements, or all communication benefits. Evidence limitations are listed separately; M is not reduced as a confidence adjustment based on the number of sources.

attribution rationaleShannon, Gallager, and Berlekamp are credited as co-authors on both parts; central symmetry is a reasoned responsibility judgment, not mechanically dividing the whole achievement by three. Fano's incompletely proven sphere-packing form and Elias's list decoding are specific precursors, retaining PERSON shares; general random-coding upper bounds are only for comparison, not separately allocated. Target Shannon a=0.19.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

M rationaleAdopted independent M3.6: Public design linking representation, move generation, evaluation, and search control under computational limits, an important design method in a professional subfield; independent institutional history confirms its theoretical exploration position. Does not assume broad industry system results or later victories, below level 4. Design complexity itself is not value; what is adopted is a judgment on the scope of this reusable professional knowledge, not claiming to measure 3.6.

attribution rationaleUsing corrected specific input budget: Shannon completed full-chain integration and programming, dominant; De G's evaluation data/experiments and Fine's strategic insights actually support search branching, vN-M prior history is a clear shared input and not split individually. Turing is parallel only, no share; Wiener appears only in reference table, not automatically credited, unverified positive inputs leave .05 unassigned. Numbers are mechanism and scope judgments, not authorship equal division, citation counts, or statistical percentages.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1955

summaryThe Simultaneous Synthesis of s Switching Functions of n Variables

M rationaleA limited mathematical extension for multi-output synthesis, tentatively M=3.5; does not double-count the 1938 framework and prior synthesis theory.

attribution rationaleTentatively allocate Shannon 75%, 25% unallocated. This is a judgment awaiting review; sole authorship does not prove causal share.

overlap boundaryIndependent increment over CAT050 awaits verification against the original text; if actually duplicate, should be revised on the same outcome.

Evidence gapsCollected works pages 780–783, or complete scan of 1955-6-8 memo MM55-114-30 (including figures).

Record year1949

M rationaleAcross generations, still an important public method in the professional subfield of graph edge-colouring; above local single observations, not confirmed as attributable to large industry or broad populations, kept between anchors 3 and 4. Not set by fame, citation count, or expected total score.

attribution rationaleShannon provided the full proposition, sharp example, odd-degree induction, and unified public argument, taking major share; Foster is a documented even-degree proof input, Petersen is an indispensable decomposition basis for that proof, each retaining .15. Remaining .05 unallocated for unidentified support. Numbers are normative responsibility estimates with causal reasons, not directly measured percentages; not averaging the two degree cases.

overlap boundarySharp upper bound floor(3Δ/2) determined by maximum degree of loopless finite multigraphs, examples achieving the bound, and original constructive proof; directly adopted as research object by later work. Excludes general graph theory, Vizing bound, and later faster or distributed algorithms.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

summaryA published set of lower bounds from counting call configurations, comparing absolute minimum state memory with separate per-call memory, and giving theoretical constructions plus the reasons absolute minima are impractical.

M rationaleDelivers state-memory lower bounds via call-configuration counting, compares absolute minimum with per-call memory, includes constructions and reasons for impracticality; an important design bound in a professional subfield. 3.3 reflects sustained structural value for professional design methods; outcome boundary is narrow, and there is a clear gap between absolute minimum and practical design, placing it mid-M3. Boundary covers only call correspondence and specified calling-party control, not supervision, statistical margins, durability, or contact load; these applicability limits restrict it from being a broader exchange-system framework.

attribution rationaleShannon takes counting, bounds, and main constructions; confirmed suggestion team 20%, Bell research environment 10%.

overlap boundaryAn applied design bound for exchange state memory, not duplicating general information theory, 1938 switching algebra, or actual deployment benefits.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1943

summaryThe memorandum gives a continuous, band- and amplitude-limited analogue of the Vernam system and argues perfect secrecy through independently uniform sampled key values; it also credits J. R. Pierce with the circular-addition implementation and saw-tooth circuit.

M rationaleThe memorandum provides a continuous system limited in bandwidth and amplitude, together with an argument for perfect secrecy under independently uniform sampled keys, yielding a specific analogue-communication construction distinct from general secrecy theory. 3.4 depends on the closure of the construction and secrecy argument: it has structural value for professional system design, but does not reach a broad platform or a field-changing result. The bounded system is confined to a bandwidth- and amplitude-limited continuous model and independently uniform sampled-key conditions; it has not been extended into a complete secrecy framework for broader channels or key processes.

attribution rationaleShannon receives 55% for the continuous analogue and secrecy argument; Pierce 25% for the specific circuit, Vernam-system prior work 15%, and the Bell environment 5%.

overlap boundaryAn independent end state of the continuous system and its implementation; it does not duplicate the general cryptosystem mathematics of CAT25, and Pierce’s circuit is not attributed to Shannon.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleThe specific reliability construction and scale bounds exceed a single technique and define a clear problem class, but are insufficient to establish a broader international methodological platform; central M = 3.7.

attribution rationaleMoore and Shannon each receive 37%; prior work on the von Neumann problem 16%, the McCulloch/Pitts model pool 3%, and unknown 7%. The individual division of proof work is unknown; the shares are normative judgments.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1951

M rationaleM3.2: by connecting the rank distribution of successive human guesses with finite-context entropy bounds, it provides a professional measurement method that is continuously reusable across institutions and therefore exceeds an isolated sample estimate. Its demonstrated benefit remains conditional measurement of printed English, placing the center at the lower end of sustained professional methods; it is not a benefit concerning constant language entropy, semantic understanding, or all compression technology.

attribution rationaleShannon designed and derived it; unnamed subjects actually made the guesses, while statistical tables and book samples supplied inputs. Unknown identities and genuinely unknown contributions are accounted for separately.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleM4.3: general and lasting distinction between random input and noncomputable information in computational capability, with independent theorem-level reuse; broader than a narrow counting technique, below a whole computing foundation or universal machine theory. 4.0 and 4.6 represent narrower-theorem versus wider conceptual-method readings.

attribution rationaleChurch’s attribution share in this outcome is 1.00%. Turing’s universal machine and computability framework are identifiable specific premises in the paper; the joint proof work of De Leeuw, Moore, and Shapiro and other mathematical inputs are listed separately, and Turing cannot be listed as a paper author.

overlap boundaryOne 1956 joint knowledge family, Theorems 1–5 and their stated effective supplements: p-machine enumeration/ordered-sequence relation to deterministic input carrying p; computable p does not enlarge this class; noncomputable p can. c-s-machine computable transition/output functions reduce to fair-coin machines in the specified eventual-output sense.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

summaryJoint published theorem and proof that equivalent resistance of a passive resistor network is jointly concave in its element resistances, yielding the common-shaft linear-potentiometer impossibility/corollary and a Hessian statement where derivatives exist.

M rationaleThe delivered proof establishes that the equivalent resistance of a fixed-topology passive resistor network is jointly concave in component resistances, and derives a potentiometer impossibility result and a Hessian conclusion where differentiable; it is a reusable network-theory result. 3.6 rests on a general theorem valid for multiple networks of the same kind and with direct design-exclusion force; the object remains fixed-topology passive networks, keeping the center below M4. The domain is limited to nonnegative passive elements and fixed topology; it has not been extended into a general framework covering active, time-varying, or topology-changing networks, nor can it be combined with concavity of channel entropy.

attribution rationaleThere is no evidence of a division of work between the authors: the two coauthors split 75% equally; each of the three explicitly cited predecessors in convexity research receives 7.5%, and the institution 2.5%.

overlap boundaryConcavity of resistor networks and concavity of channel entropy are distinct end states; only the fixed-topology passive-network theorem and corollary are counted.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

summaryJoint paper describes computational assistance for switching-circuit design, including reduction/selection procedures and machine implementation; it is more than a general forecast but its relation to earlier Boolean switching and later synthesis must be resolved.

M rationaleUnder the independent-delivery condition, the paper realizes reduction and selection procedures for switching circuits by machine, exceeding a forecast and sufficient to constitute an important computer-assisted design method in a professional subfield. 3.5 is an estimate conditional on “the implementation having an independent endpoint from existing methods”; its executability is strong, but the independent endpoint and coverage boundary are not closed; it is not an interval midpoint. The independent increment over earlier Boolean switching methods and subsequent synthesis work is not closed, and the program covers only reduction and selection of switching circuits; machine implementation cannot be expanded from this into a broader field change.

attribution rationaleThe division of coauthor work is unknown: Shannon and Moore each receive 40%; prior switching theory and machine resources each receive 10%.

overlap boundaryCAT80 is machine-assisted reduction/selection design, independent of the 1938 Boolean mapping, CAT50 general synthesis, and CAT83 specific tight bound.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1980

M rationaleM = 2.8: proof and measurement treatment of uniform throwing-and-catching timing and pattern constraints constitute a narrow but complete professional method; it is not extrapolated into general movement science or apparatus impact.

attribution rationaleShannon authored the uniform-timing proof and preliminary measurement treatment; because apparatus/recording assistance is not separated, 15% is reserved at the center, and the entire mathematical–measurement treatment is counted as only one end state.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

M rationaleM2.8: Theseus integrates trial-and-error search, route retention, and re-finding under changed conditions into an observable, repeatable physical learning demonstration; it has sustained professional explanatory value and approaches M3, but the delivery covers one limited maze device and does not expand into general AI or a new learning theory.

attribution rationaleShannon was responsible for the strategy, mechanical design, and control integration; Betty Shannon actually wired the mouse body and relays, with unnamed implementation and prior work separately retained. The central 72%/23%/5% reflects these documented actions, not an authorship rule.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1942

M rationaleM3.5: durable professional-subfield enumeration and complexity method with concrete independent later use; broader than one teaching example or alternative proof alone, narrower than international foundational computing capability. 3.0/3.7 vary this scope-weight judgment; not inferred from target score.

attribution rationaleJoint delivery and independent technical attribution support symmetric research responsibility. A .62 joint pool is retained after explicitly allocating credit to MacMahon’s starting problem/identity, Euler and Gupta’s methods, Goeltz computation and narrower formula/asymptotic inputs; each author gets .31 by source-symmetric judgment, not by author-count alone. There is no proof of an internal lead; .44/.24 mirrors are uncertainty alternatives, not assertions. Shannon’s 1938 application setup is within his personal row, not a new external charge or a duplicated 1938 outcome.

overlap boundaryOne 1942 joint knowledge family: proof of the MacMahon generating identity; enumeration and numerical recurrence schemes/table; rigorously derived comparison bounds and an almost-all switching-function series-parallel-realization lower bound. Their semi-empirical approximation is labelled as such; no exact asymptotic formula was found.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1948

M rationaleM4.1 corresponds to an important knowledge result for a major industry/professional field: the 1948 paper organizes sampling, quantization, coding, regeneration, reconstruction, and bandwidth–power–noise trade-offs into a rigorous public analysis; independent technical history identifies it as key work consolidating PCM theory. It does not own the invention or deployment of PCM, or all benefits of digital communication.

attribution rationaleOliver, Pierce, and Shannon jointly completed this paper, but available sources do not distinguish their specific divisions of work, so direct attribution to the three uses a symmetric point estimate. Reeves’s conception and circuit work, and relevant work by Goodall, Grieg, Black, Edson, Clavier, Panter, Meacham, and Peterson, are separately listed in the same complete budget. Nyquist and Hartley remain in the non-exclusive prior-work pool for sampling and harmonic methods; the historical sequence is still insufficient to establish their respective shares or direct lines of inheritance, so no separate individual shares for the two are adopted. There is as yet no evidence that the Bell Labs project had an independent marginal role beyond named actions; the corresponding unallocated share is not assigned to Shannon or other authors. Shannon’s existing capacity theory is only an input to this item and is not separately accounted for again.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

M rationaleM = 2.7: what is public is a reusable simplification strategy for a two-choice human–machine game (behavioral memory, random fallback, and scoring); the original machine, deployment, and broader field impact are not counted.

attribution rationaleThe text makes clear that this is a simplification based on the Hagelbarger machine: Shannon’s memory/fallback strategy and explanation are the added delivery, while the original machine is a substantive antecedent input; the central accounting recognizes modification and does not call it creation from scratch.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

M rationaleM3.5 corresponds to a sustained, reusable methodological bridge within a professional subfield: the original work completes a simplified physical derivation and implementable construction, and Kalman later explicitly takes up its white-noise–linear-system representation; it is not Wiener/Kolmogorov theory itself, and there is no evidence covering all prediction and filtering applications.

attribution rationaleSources do not distinguish the internal tasks of Bode and Shannon, so the center is strictly symmetric and two complete asymmetric mirrors are preserved. Wiener, Kolmogorov, conventional Fourier/circuit/spectral techniques, and Hoel are all actual inputs in the text. Kalman only proves subsequent uptake; his state-transition, nonstationarity, and growing-memory results are another outcome and are not in this budget.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

summaryA published worked exposition/construction of signed-digit representations, odd-radix sign-symmetric arithmetic and rounding, and an even-radix half-integer digit extension with an explicitly inconvenient infinite representation of integers.

M rationaleThe delivered system presents signed digits, odd-radix symmetric arithmetic and rounding, and extends them to even-radix half-integer digits while explaining their defect; it is a complete but limited mathematical construction and explanatory outcome. 2.4 is jointly determined by the construction’s completeness and range of application: stronger than local instructional help, but still insufficient to count as a lasting structural change in a professional subfield. Originality for all balanced numeral systems is not established; the even-radix extension also produces infinite representations of integers, limiting applicability, and symmetric arithmetic and rounding remain one outcome family.

attribution rationale80% goes to the systematic construction and explanation; unnamed prior work on representations and other inputs retain 12% and 8%, respectively.

overlap boundaryOnly the signed-digit representation and symmetric-arithmetic family are counted; computer deployment and the invention of all numeral systems are not counted.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

summaryA published proof that simultaneous realization of the 16 Boolean functions of two variables needs at least 18 contacts, paired with an 18-contact construction credited by the author to an unnamed MIT switching-course student.

M rationaleThe 18-contact lower-bound proof and 18-contact construction jointly close a clear necessary-and-sufficient minimality result, forming a reviewable professional limit for switching-circuit design. 2.7 is placed between a small institutional-scale result and an important professional-subfield result because endpoint closure is high but object scale and extension are limited. This is a specific minimality conclusion for the 16 Boolean functions of two variables; no general synthesis theory extendable to arbitrary variables or function families has been established, and its object and scale of application are limited.

attribution rationaleOne necessary-plus-sufficient end state: Shannon’s proof 55%, the unnamed student’s construction 30%, prior work 10%, and the course environment 5%.

overlap boundaryIndependent of the 1938 general Boolean switching analysis: this case is the tight 16-function bound of 18 contacts, and it also differs from the CAT50 synthesis method.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1946

M rationaleIt gives an asymptotic transient classification for chains of unilateral four-terminal linear networks and a necessary/sufficient argument for bounded no-net-gain conditions; M = 3.0, limited to this specialized network class, with hardware, adoption, and broad communications impact not counted.

attribution rationaleJoint authorship and signatures support a neutral center of 35% each for Shannon and Dolph; MacColl 12%, Wise 3%, Hamy and Winter 2% each, and unseparated methods/laboratory/transmission 11%. These are not observed labor percentages.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

M rationaleM2.0: a complete physical embodiment of four-operation Roman-numeral arithmetic makes an unusual numerical representation into an operable demonstration, reaching the bounded concrete benefit of M2; the new combination of control and representation did not change general computing capability, and available evidence supports this device rather than an industry system.

attribution rationaleShannon designed the Roman encoding, relay circuits, and whole machine; the designers of the Marchant desk calculator supplied the control model, while unnamed implementation and general electronic inputs are separately retained. The central 85%/8%/7% is normative accounting.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleThere is a durable, verifiable alternative proof in an important professional subfield; other proofs already exist, so it cannot rise to M5/6 by borrowing the broad impact of the entire theorem or network optimization; the center is M = 3.

attribution rationaleThe specific proof was jointly delivered by three authors, 23% each; prior-work results and conservation inputs 21%, institutional support 3%, and unknown 7%. Equal shares only indicate no discernible individual difference under current evidence.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

M rationaleThe public patent gives a complete conversion chain of equal-step charge reduction, binary counting, transmission, and decoding; M = 2.5 counts only reusable component design, not hardware performance, deployment, PCM priority, or later benefits of digital communication.

attribution rationaleTwo named inventors jointly disclosed it; current records cannot distinguish internal contributions, so the center gives Shannon and Oliver 40% each, with the remaining 20% retained for unseparated prior circuit/counting methods, institutional work, and publication work; this is not a mechanical allocation by number of names.

Evidence gapsThis item already has a frozen review chain; this time no public specific full-text URL/page number was extracted from the corresponding review artifact, and a catalog link cannot substitute for it. This is not a conclusion that the original text is missing.

Record year1982

M rationaleM2.2: three sets of mechanisms and phase control develop a static record display into a continuous watchable mechanical simulation; both professional production demonstration and viewing benefit belong to the delivered limited result, so it is above an isolated small technique and in the lower part of M2. A visual simulation fixed by support rods is not counted as a genuine throwing-and-catching robot or a record of human juggling.

attribution rationaleShannon completed the design and integration of the mechanical scene; Betty supplied the triggering toy, Phil Stone and family participated in construction, and Dennis Soldati supplied the selected record values. Attribution retains these explicit collaborative roles and does not turn it into a solo work.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

summaryPublished PIRE survey organizes non-numerical computation around brain/computer comparison, Turing machines, learning machines, game machines and self-reproducing automata, with bibliography and research questions.

M rationaleThe public survey organizes non-numerical computation around themes including brain–computer comparison, learning, games, and self-reproducing automata, with bibliography and research questions; what is counted is bounded explanation and research navigation. 1.6 is supported by multi-theme synthesis and professional dissemination value, above a short commentary, but no lasting structural improvement for a small institution or professional subfield has yet been established. There is no evidence of a new technical endpoint, adoption range, or behavioral change; the theories, machines, and projects cited must not be accounted for again.

attribution rationaleAllocate only explanatory work and research-agenda organization: Shannon 65%, cited research content 25%, PIRE editorial dissemination 10%.

overlap boundaryDo not count technical results of Turing, von Neumann, and others again, nor duplicate Shannon’s existing automata/game end-states.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleWhat is public is a closed-loop geometric optimization for finite memoryless discrete-channel capacity: convex-hull–entropy-surface representation, deletion of interior points, binary-output extreme inputs, and a maximum-vertical-difference construction. M=1.5; it does not duplicate general information theory or unproven industrial adoption.

attribution rationaleShannon completed the geometric integration, derivation, and binary-output construction, central 70%; Muroga’s prior capacity results 14%, and Alexandrov’s and Hopf’s convex-body tool inputs 8% each, are normative estimates.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1956

M rationaleM2.2: the demonstrable delivery is the joint selection of an early interdisciplinary collection, its three-part framework, and reader orientation, with continuing organizational value at the professional-community level; independent sources only corroborate joint-editor status and cannot upgrade it into technical invention of a field or grand impact.

attribution rationaleDirect evidence for the editorial division between Shannon and McCarthy is lacking, so this item provisionally assigns 25% personal attribution to each. The authors’ contributed papers listed in the preface, Princeton Logistics Project and ONR support, Tucker, the publishing team, Julia Robinson’s review, and E. Powanda’s secretarial service are all included in one complete budget. The Princeton project and ONR funding are treated as one support chain; contributors participate in the editorial outcome only through provision of papers and do not transfer technical contribution of their papers; editors’ already-counted personal actions are not claimed again in contributor or prior-work pools.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

M rationaleThe implemented dedicated relay synthesizer uses an existing four-variable catalogue, paper-tape search, equivalence transformations, and shortcuts, and can repeatedly return low-contact designs; M=1.8. Scope is limited to four variables/two terminals and cannot be generalized into universal automated design.

attribution rationaleJoint-completion baseline is Shannon 35%, Moore 35%; the 402-class catalogue already attributed to Moore takes 15%, and 15% is unseparated implementation/equipment/existing logic, yielding central Shannon 35%, Moore 50%, unknown 15%.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1950

summaryThree-page memorandum gives a floor-strip/brush/rectifier arrangement intended to supply power or signalling to a freely oriented moving vehicle without trailing wire or trolley, including geometry margins and variants.

M rationaleThe three-page memorandum proposes a combination of floor strips, brush contacts, and rectifiers, gives geometric margins and variants for power/signalling to freely oriented vehicles, and delivers a specific engineering scheme. 0.65 is supported by specificity sufficient for local engineering reference, but key operating conditions remain unresolved, so it is only in the substantial-local-help band. The scheme closes only the geometric and circuit conception for power or signalling; moving contact, wear, electrical load, and short-circuit margins are not closed, and it cannot be incorporated into later wireless-power lineages.

attribution rationaleThe public design itself is deliverable: Shannon 85%, existing component technology 5%, Bell environment 10%; no prototype or commercialization is required.

overlap boundaryCounts only the wired-floor power/signalling-device design; it does not infer wireless power, a patent, prototype, or commercial outcome.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1949

summaryPublished review introduces Wiener’s cybernetics programme, summarizes its statistical/control and nervous-system arguments, identifies specific editorial/technical reservations, and notes his pessimism about adaptation to automation.

M rationaleThe review summarizes control, statistical, and nervous-system arguments and makes specific editorial and technical reservations, while also noting concern about adaptation to automation; it is a professional explanatory delivery with independent judgment. 0.85 is supported by its combined explanatory and critical functions, below the important-benefit anchor over many years; lacking continuing effects, it is not rated M1 or above. The original theory and impact of automation are not assigned to this review, and no evidence shows it changed research or policy.

attribution rationaleAssess only the book-review service: Shannon’s summary, discrimination, and criticism take .55; the reviewed original book and technical content are a .30 source-input pool, without making Wiener a co-owner of the review; PIRE editorial-publication service .15.

overlap boundaryCybernetics technology and automation effects remain with Wiener’s original outcomes; this item counts only Shannon’s introduction, discrimination, and criticism, and the source-input pool does not propagate to Wiener’s person ledger. This round only withdraws personal attribution to the reviewed author as a producer of a derived service; the source-content pool does not establish cross-outcome deduplication. M assesses only the added editorial/explanatory service here, and does not claim the original work’s knowledge and continuing effects again.

Evidence gapsUses existing sources and limited-major-family assessment; it does not exhaust continuing use, independent alternative paths, or source inputs. This round did not reread the original work in full.

Record year1956

summaryOne-page published editorial translates a technical boundary into a public methodological warning: concepts do not transfer to psychology/economics merely by word substitution and require hypothesis/experiment.

M rationaleThe one-page editorial turns a technical applicability boundary into a clear methodological warning, requiring hypotheses and experiments when crossing into psychology or economics; it is a limited but explicit public correction. 0.55 is supported by a clear warning usable for one methodological judgment, but lacks evidence of continuing adoption and lies in the middle of the substantial-local-help band. There is no new theorem, evidence of changed external behavior, or cross-disciplinary success, and its length and argumentative scope cannot support institutional-level long-term improvement.

attribution rationaleThis is authorial commentary rather than a new theorem: Shannon’s argument and wording take 90%, editorial dissemination 10%.

overlap boundaryIndependent bounded-commentary end-state; it neither duplicates information-theory technical credit nor claims to change the behavior of external disciplines.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1949

summaryPublished PIRE review explains Kiss’s three-part lattice/generalized-Boolean programme and qualifies possible switching/nerve-network applications.

M rationaleThe review explains a three-part lattice-theory and generalized-Boolean programme and cautiously qualifies switching and neural-network applications; it counts a structured explanation and boundary clarification for readers. 0.75 is supported by structured explanation and cautious qualification; length, audience, and continuity do not establish M1 or above.

attribution rationaleAssess only the book-review service: Shannon’s explanation and judgment take .55; the reviewed work and theoretical content are a .30 source-input pool, without making Kiss a co-owner of the review; PIRE editorial-publication service .15.

overlap boundaryDo not allocate credit for CAT62 information-lattice theorems or assign the reviewed work’s technical achievements to Shannon; the source-input pool does not propagate to Kiss’s person ledger. This round only withdraws personal attribution to the reviewed author as a producer of a derived service; the source-content pool does not establish cross-outcome deduplication. M assesses only the added editorial/explanatory service here, and does not claim the original work’s knowledge and continuing effects again.

Evidence gapsUses existing sources and limited-major-family assessment; it does not exhaust continuing use, independent alternative paths, or source inputs. This round did not reread the original work in full.

summaryA narrow behavioral candidate already included in the historical ledger: MIT research-seminar teaching. This revision retains the responsibility point; public sources remain unresolved and it does not claim re-verification.

M rationaleReason given by the historical assessment record (this revision did not find original links readable and verifiable item by item): participant testimony confirms frequent seminars and a complete course with new content in every lecture, establishing positive repeated teaching; .3 reflects a full course and frequent activity beyond a single lecture. Current sources do not determine cross-year continuity, so M1 strong continuing results and 1.4 cannot be inferred merely from “frequent” and new results. New scientific content is not claimed again in this teaching family.

attribution rationaleParticipant testimony confirms Shannon delivered seminars and a complete course; MIT facilities and delivery support retain a small share, while Gallager is a witness and students are beneficiaries; neither automatically receives a production share.

overlap boundaryOnly this narrow teaching/delivery/referral behavior; it does not duplicate paper content, communication infrastructure, or subsequent outcomes of taught/referred people.

Evidence gapsThis revision’s search of existing Shannon evidence items still found no readable public URL and page numbers for itemized verification; the public-source gap continues; the historical READY label does not equal original-text verification in this revision.

Record year1950

summaryPublished review describes the Harvard Mark II book’s relay-calculator organization, operation figures, programming/checking themes, and reference value.

M rationaleThe review summarizes relay-calculator organization, operational data, programming and checking themes, and judges reference value, constituting explicit but limited navigation for professional readers. 0.35 is supported by coverage of specific themes and value in reference selection; equipment and technical outcomes are not assigned to the reviewer.

attribution rationaleAssess only the book-review service: Shannon’s summarizing, selection, and judgment take .55; the manual, equipment, and technical content produced by Aiken and the Mark II team are combined as a .35 source-input pool, not as review co-owners; PIRE editorial-publication service .10.

overlap boundaryEquipment, organization, and programming outcomes remain with the Harvard team’s original outcomes; this item only counts navigation for professional readers, and the source-input pool does not propagate to Aiken’s person ledger. This round only withdraws personal attribution to the reviewed author as a producer of a derived service; the source-content pool does not establish cross-outcome deduplication. M assesses only the added editorial/explanatory service here, and does not claim the original work’s knowledge and continuing effects again.

Evidence gapsUses existing sources and limited-major-family assessment; it does not exhaust continuing use, independent alternative paths, or source inputs. This round did not reread the original work in full.

Record year1950

summaryScientific American article explains chess-position coding, evaluation, selective search, subprogram architecture, randomness and proposed learning limits for a chess-playing computer.

M rationaleConnects position coding, evaluation, selective search, program architecture, randomness, and learning limits into a publicly readable design explanation, giving systematic local-understanding help M0.2; underlying design and master experience remain inputs, and it adds no implemented machine or chess-strength outcome.

attribution rationaleAssess only the public-explanation service: Shannon’s explanatory organization takes .75; De Groot’s research and Torres’s existing chess machine are combined as a non-person source-input pool .15, without making the two original authors co-owners of this explanation; Scientific American editorial-publication service .10.

overlap boundaryAll underlying chess-program design remains in CAT54 original outcomes; this item counts only independent public explanation, and the source-input pool does not propagate to De Groot or Torres person ledgers. This round only withdraws personal attribution to the reviewed author as a producer of a derived service; the source-content pool does not establish cross-outcome deduplication. M assesses only the added editorial/explanatory service here, and does not claim the original work’s knowledge and continuing effects again.

Evidence gapsUses existing sources and limited-major-family assessment; it does not exhaust continuing use, independent alternative paths, or source inputs. This round did not reread the original work in full.

Record year1968

summaryEncyclopaedia Britannica article introduces the five-part communication model, entropy/coding/noise, bandwidth tradeoffs, filtering/prediction and stated application boundaries with bibliography.

M rationaleThe five-part communication model, entropy coding noise, bandwidth tradeoffs, filtering prediction, and application boundaries form a complete reference explanation; .2 reflects connected multi-level explanation. CAT37 technical discoveries and overlapping explanatory dissemination are excluded. Encyclopaedia publication and reprinting confirm availability but cannot themselves establish strong continuing group results above 1.

attribution rationaleThe same actual explanation/reference-resource boundary; original research inputs remain only causes of that explanation and are not counted again as new discoveries. Post hoc checks retain every role and unknown remainder in the adopted budget, and do not transfer removed M or old shares to the individual. This is not a fresh independent a reassessment of all 76.

overlap boundaryCAT37 retains all technical-theory credit; this case counts only independent encyclopaedia explanation and does not duplicate similar dissemination in CAT60/61/65.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1954

summaryJoint teaching design specifies a relay plugboard, manual/experiment sequence and named circuits; it reports a constructed 24-relay experimental model and demonstration to Digital Control Techniques professors.

M rationaleThe 24-relay model, manual/experiment sequence, and named circuits were actually constructed and demonstrated, delivering an operable teaching toolkit; .3 indicates more complete local operational help than a single explanation. Multi-year courses or institutional structural improvement are unproven, so reusability of design cannot yield 1.9; graduate outcomes are not required to recognize the tool’s value.

attribution rationaleThe same accepted knowledge/service end-state and the same causally operative action scope; this revision only corrects the M-scale rationale and does not make the outcome an invention/student consequence. Retain the entire global budget and do not reallocate it to the individual. Shares are normative judgments, not measured proportions; this is not a 76-item independent a re-review.

overlap boundaryThis is the end-state of teaching apparatus and experiment sequence; it does not duplicate CAT80 machine-aided design or CAT83 contact tight bounds.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

Record year1953

M rationaleA four-page popular article explains programmed computers, examples, limitations, and possible social consequences, delivering one explicit substantive knowledge explanation; take .1 substantial local help. Anticipated effects and cited computer technology are not counted, and there is no fact of 1.3 strong continuing group benefit.

attribution rationaleThe same actual explanation/reference-resource boundary; original research inputs remain only causes of that explanation and are not counted again as new discoveries. Post hoc checks retain every role and unknown remainder in the adopted budget, and do not transfer removed M or old shares to the individual. This is not a fresh independent a reassessment of all 76.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

summaryA narrow behavioral candidate already included in the historical ledger: newspaper and telegram delivery labor. This revision retains the responsibility point; public sources remain unresolved and it does not claim re-verification.

M rationaleReason given by the historical assessment record (this revision did not find original links readable and verifiable item by item): M=0.08: counts only verified personal newspaper and telegram delivery labor, without incorporating communication infrastructure or beneficiaries’ outcomes.

attribution rationaleArchive applications and independent obituaries confirm actual delivery; the central 90% emphasizes Shannon’s direct labor, with 10% retained for delivery infrastructure, and beneficiaries are not listed as producers.

overlap boundaryOnly this narrow teaching/delivery/referral behavior; it does not duplicate paper content, communication infrastructure, or subsequent outcomes of taught/referred people.

Evidence gapsThis revision’s search of existing Shannon evidence items still found no readable public URL and page numbers for itemized verification; the public-source gap continues; the historical READY label does not equal original-text verification in this revision.

Record year1958

summaryPublished BAMS memorial/synthesis maps von Neumann’s automata work into reliability, self-reproduction, coding/language and unfinished research questions, with selected bibliography.

M rationaleOrganizes existing automata work into reliability, self-reproduction, language coding, unfinished questions, and bibliography, giving actual readers local technical navigation M0.1; scientific results belong to the original researchers, and the memorial article itself does not establish stronger continuing benefit.

attribution rationaleAssess only the synthesis-review service: Shannon’s organization, explanation, and research navigation take .60; von Neumann and other automata technical content are combined as a non-person source-input pool .35, without making the reviewed author a co-owner; BAMS editorial-publication service .05.

overlap boundaryAll automata technical achievements remain with von Neumann and collaborators’ original outcomes; this item counts only Shannon’s synthesis review, and the source-input pool does not propagate to von Neumann’s person ledger. This round only withdraws personal attribution to the reviewed author as a producer of a derived service; the source-content pool does not establish cross-outcome deduplication. M assesses only the added editorial/explanatory service here, and does not claim the original work’s knowledge and continuing effects again.

Evidence gapsUses existing sources and limited-major-family assessment; it does not exhaust continuing use, independent alternative paths, or source inputs. This round did not reread the original work in full.

Record year1963

summaryThe 1962 Rice lecture/published proceedings provides a historical scaffold for information processing, explains bit/capacity, surveys named projects, and frames automation’s distributional problem.

M rationaleIt taught and published the history of information processing, bit/capacity, and automation’s distributional problem; take .1 for one substantive local knowledge explanation. Social risk remains as discussion; unrealized predictions are not treated as social benefit or harm, and there is no multi-year teaching/institutional change to support the old 1.7.

attribution rationaleThe same accepted knowledge/service end-state and the same causally operative action scope; this revision only corrects the M-scale rationale and does not make the outcome an invention/student consequence. Retain the entire global budget and do not reallocate it to the individual. Shares are normative judgments, not measured proportions; this is not a 76-item independent a re-review.

overlap boundaryDo not turn examples such as Samuel, Slagle, Wang, and von Neumann into Shannon technical outcomes; risk discussion is not measured harm.

Evidence gapsExisting sources and a limited set of principal families are retained for the assessment; evidence of continued use, independent alternative paths, and predecessor contributions is not exhaustive. This revision did not reread the full texts of all original works.

summaryA narrow behavioral candidate already included in the historical ledger: recommending Kelly’s paper to Thorp. This revision retains the responsibility point; public sources remain unresolved and it does not claim re-verification.

M rationaleReason given by the historical assessment record (this revision did not find original links readable and verifiable item by item): M=0.04: counts only one verified referral of a technical paper, without counting the referred paper or subsequent transformation again.

attribution rationaleThe recipient’s recollection confirms Shannon personally referred an existing paper; Kelly is the paper’s author and Thorp its subsequent transformer, but neither is a co-producer of the referral, so this narrow end-state belongs to Shannon.

overlap boundaryOnly this narrow teaching/delivery/referral behavior; it does not duplicate paper content, communication infrastructure, or subsequent outcomes of taught/referred people.

Evidence gapsThis revision’s search of existing Shannon evidence items still found no readable public URL and page numbers for itemized verification; the public-source gap continues; the historical READY label does not equal original-text verification in this revision.