Volume 19 of W. Ross Ashby's Journal
is loading...

Warning: Your browser does not have JavaScript enabledJavaScript has been used to implement many features on this Web site. Unless you enable JavaScript, the following things will not work for you: pop-up image previews, feedback links, random page links, the jump-to-any-page, jump-to-any-volume boxes, and much more. It will annoy you...

1954



Volume 19



4993+03	4993+04

Summary: Algebraic form of "immediate effect".
Summary: How to deduce the connexions from input to variable. 5055 Diagram of immediate effects (D.I.E.) and parameters Input deducing connexions
4994	4995

Summary: "Connexion" depends on the operations brought to its study. Connexion meaning of Diagram of immediate effects (D.I.E.) deducing

4996	4997

Isomorphism electromechanical example Higher geometry of fields and matrix theory [17]: An electromechanical analogy (spring and mass / condenser etc) in detail 4998.
Summary: An exact electromechanical analogy.
4998	4999

Summary: Isomorphism. Isomorphism formal definition Machine isomorphism defined
Summary: Homomorphism. Homomorphism example
5000	5001

Summary: A problem.
Constraint machine as Machine as constraint Epistemology [48]: Deducing the machine from the behaviour 5003.
5002	5003

Summary: What is a machine? 5056
Summary: Game that is wholly arbitrary. 5024 Continuity in games Games wholly arbitrary
5004	5005

Design of reducible machine Reducibility and design
Joining design in
5006	5007



5008	5009

Summary: Design of a machine part by part. 5474, 5072 Triunique example
Summary: Example of triunique relation. Feedback bit by bit
5010	5011

Summary: Behaviour when the feedback can carry only one bit.
Regulation all types collected
5012	5013



5014	5015



5016	5017



5018	5019


Summary: The relations between regulators are complex and hardly worth developing.
5020	5021

Problems solving problems Solution (to a problem) finding
Summary: Problems that are not wholly new. 5066 Summary: A numerical illustration of continuity as a constant. 4569 5005 4597 Constraint continuity as Continuity as constraint
5022	5023


Summary: Example of the effect of continuity as a constraint. Caution: See 5054. 5636 Diagram of immediate effects (D.I.E.) and memory Isomorphism in memory Memory isomporphism in
5024	5025

Summary: Effects of channel capacity in joining two systems, only one of which is observed. Epistemology [49]: As the coupling between B and A is made richer, either by increasing channel capacity or by adding immediate effects, so will what is in B affect x the sooner. 5026. DIAGRAM
Entropy during search Hunt and stick information flow Information during hunt and stick Searching information flow Selection informatiom flow Trial and error gives information
5026	5027

J - function (Ashby) blemish

5028	5029



5030	5031



5032	5033



5034	5035



5036	5037

Summary: Flow of information during trials.
Turing machine principle
5038	5039



5040	5041

Summary: The Turing machine.
Machine reproduction of Reproduction machines for
5042	5043


Summary: A machine that makes itself. 5088
5044	5045

Hunt and stick in logic machine Logic machine Strategy of trial and error Veto in equilibria Veto use by McCallum
Summary: McCallum and Smith also use the method of veto. 5583 Requisite Variety, Law of in set theory
5046	5047

Summary: Requisite variety stated algebraically. 5075 Isomorphism in memory
Memory isomporphism in
5048	5049

Summary: The field, when memory is used, is not isomorphic with that found when all is observed. Epistemology [50]: If only components in J are observable, and we know that the J-states have been, in succession, a₀,a₁, ... ,a_k, and if the machine's mapping is S, then the machine must now be in one of the states in EXPRESSION. 5050.
Trajectory deducing others Higher geometry of fields and matrix theory [6]: Knowledge of one form EXPRESSION for all values of xº₁...xº_n;t is sufficient to define all the others. 5051.
5050	5051



5052	5053

1955

Summary: Studying a system via one variable. Continuity in discrete sets Epistemology [51]: The whole system can be studied if one variable is observable; in particular, all the canonical equations can be deduced 5054.
Summary: (Comment on "continuity") Input found by testing
5054	5055

Summary: Studying a system... Constraint machine as Machine as constraint Protocol deducing input
Summary: "Machine" as constraint in a sequence. But see 5074
5056	5057

Absolute system algebraic test Canonical equations as [TPT^-1] Canonical equations in set theory Operator from protocol Protocol deducing absolute system

5058	5059


Summary: Converting observed behaviour to specification of machine. 5063, 5081, 6071 Summary: Beware! The "design" of a machine costs simply what is necessary to get it selected from what is available. There is no unique quantity "in" a machine. 5006 Design amount in machine
5060	5061

Summary: The quatity of design in measured by the size of the set that is drawn from. 5070, 5071, 5072 Behaviour operations with Operator for behaviour

5062	5063

Summary: Algebra can operate directly with observed behaviour. Everywhere defined mnemonic for Single-valued mnenomic for
Summary: Mnemonic for "everywhere defined" and "single valued". Summary: Metron and logon. Logon defined Metron defined
5064	5065

Summary: What is problem solving? I say it is finding an element in a set. Problems solving problems Solution (to a problem) finding Strategy of problem solving

5066	5067

Summary: Problem solving. Markov process / chain as problem solver
Summary: Another example of a self-locking system, this one harmful. Chain on wheel, as self locking system Self-locking system example of chain on wheel
5068	5069

Deputy design necessary Design amount for deputy Machine amount of design
Summary: Briefing a deputy calls for the same capacity as doing the job oneself. 5073 Summary: Amount of design required for a machine to do a job.
5070	5071

Machine amount of design
Summary: More on "the amount of design in a machine". Amplifier of regulation Regulation amplification of
5072	5073

Summary: Why build a regulator? Machine has two aspects
Summary: A machine embodies a transformation and, in addition, may act repetitively. Summary: Two problems. Homeostat theory of Unsolved problems [16]: Represent this by a pay-off matrix:- the case when the disturbance is compound and one component varies only occasionally. 5075. Unsolved problems [17]: State the thesis that "selection forces adaptation" in terms of disturbance and regulation. 5075.
5074	5075

Summary: Theory of the homeostat. Amplifier demonstration Magnet in selection-amplifier Selection demonstration
Summary: Demonstration of selection-amplifier. Prediction falsifies itself
5076	5077

Interaction and prediction Joining and prediction

5078	5079

Summary: Systems that interact finitely with their observer. Machine as constraint
Machine algebraic in protocol
5080	5081

Summary: The structure of a machine, derived from a protocol.
Protocol deducing machine
5082	5083

Summary: Structure of machine in a protocol. 5170, 5192, 6071 Equivalence relation over protocol Homomorphism over protocol Protocol equivalent relations over
Summary: The problem of "simplifying" a machine. Gödel's theorem described Simplicity in homomorphisms
5084	5085



5086	5087

Summary: Godel's theorem as process. Virus as self-reproducer
Summary: More on the self-reproducing system. 5261 Games and constraint Reproduction example (artificial) Strategy must fit constraints
5088	5089

Metric and convergence
Summary: A strategy must be related to the actual constraints of a system.
5090	5091

Capacity independent of reducibility Reducibility independent of channel capacity
Summary: Reducibility and channel capacity are independent. 5532 Linkage
5092	5093

Summary: Finding a maximising formula. Max( ) (function) strategy of
Joining tensor method Tensor coupling, re-stated
5094	5095



5096	5097

Algebra of coupled systems
Summary: Formally stated method of coupling by using the tensor product of Riguet.
5098	5099



5100	5101



5102	5103

Summary: Tensor method of coupling systems. Summary: Tensor method of coupling systems.
Absolute system examples, list of
5104	5105

Latent roots list of Matrix list of special matrices

5106	5107

Summary: List of shorthand symbols that may be found between pages 4416 and 4446, and in the Index. Shorthand list of symbols Personal notes [22]: I start shorthand, 4416, I stop it, 4446, List of symbols, 5108.
Summary: Review of Pringle. Communication involves set of acts Evolution Pringle on
5108	5109

Entropy and organisation, in cream Organisation and entropy
Summary: The problem of the glass of milk.
5110	5111

Summary: What is meant by "He does not know......." Knowing and size of set Epistemology [53]: "A does not know B's phone number" means, operationally, that if A is forced to make an attempt at calling, his action must be deternmined by some other, random, factor 5112.
Summary: With systems in general, do not look for the law; each must be studied and taken as it is found. Over a set of systems a constraint may be found. 5114.7. Society [64]: Societies far from the atomic level are too many to give a necessary law, though a constraint may be found in practice. 5113, 5115.
5112	5113

Summary: Bridgman on operations. Atom redundancy among atoms Operational principle Bridgman on Redundancy at atomic level Quotations [6]: What a man means by a term is to be found by observing what he does with it, not by what he says about it. Bridgeman 5114.
Society [64]: Societies far from the atomic level are too many to give a necessary law, though a constraint may be found in practice. 5113, 5115. DAMS (Dispersive and Multistable System) [95]: The "law" of a system (or set of systems) far from the atomic level can be found only empirically. 5115.
5114	5115

Summary: There is no "law" to be found in systems much removed from the atomic level. 5142 Laws of nature only near atomic level
Summary: On causality. (Continued over) Black box, problem of the "cause" and "why" Cause Rapoport on Information statistics as reducer Large system and statistics Operational principle "why?" and Statistic as diminisher of information Why ...? meaning of
5116	5117

Cause Rapoport on Operational principle Rapaport on Why ...? meaning of
Immediate effect equals "cause"
5118	5119



5120	5121


Input and "cause"
5122	5123


Correlation and 'cause' Memory as "cause"
5124	5125

Goal as "cause" Teleology as "cause"

5126	5127

Constraint as "cause"
Summary: Concepts of cause, why, because etc related to Black Box theory.
5128	5129

Channel capacity channels and lattice Diagram of immediate effects (D.I.E.) various, and lattice Lattice of diagrams of immediate effects
Summary: The simplifications obtained by cutting channels lie on a lattice. 5135 Reducibility in organisation
5130	5131

Organisation advantage of
Design and reducibility Regulation and reducibility
5132	5133

Summary: Why the organised and inter-connected can usually defeat the unorganised and individualistic.
Channel capacity channels and lattice
5134	5135

Summary: Quantity of organisation.
Network growing instability Society [65]: Each unit, to improve its own regulation, improves its own channels of communication. But this adds richness to the channels in the whole system, so the whole moves on to a long and complex trajectory, till it hits a critical state. 5137.
5136	5137

Absolute system in history Disintegration of society England as absolute system Independence of England Isolation of England
Beginning "natural" Initial state "natural"
5138	5139

Impulse as "starter" Stimulus as "starter"
History laws of Laws of nature in history
5140	5141

Society [66]: Toynbee's "laws" of societies 5142.

5142	5143

Summary: Application of system theory to History.
Equilibrium giving binary relation Machine binary relation in Relation binary, in machine Set or Ensemble operations and machine
5144	5145

Summary: How can a machine express a binary relation? 5181
Summary: Law of Experience algebraically. 5155, 5209.7 Experience in algebra
5146	5147

Equivalence relation over times Simplicity by fusing times Time equivalence relation time

5148	5149



5150	5151

Summary: Simplification by running together, or deleting, the elements of time. 5165, 5245
Summary: "Loss of control" in set theory. 5155 Control in set theory Convergence (of lines of behaviour) loss of control Derivative and set theory Information non-transmission
5152	5153

Summary: The conditioned reflex, and shedding of scatter by going to an equilibrium. 5249.1, 5277.8 Equilibrium as noise-destroyer Noise method for destroying
Experience rigorous form Input forcing convergence
5154	5155



5156	5157


Summary: Law of Experience, given rigorously. 5179 Definition as transformation Mathematics nature of
5158	5159

Summary: Kershner and Wilcox' book.
Canonical equations information in Information in canonical equation Input of mathematical form Mathematics variety in Variety in canonical representation
5160	5161



5162	5163

Summary: Variety in mathematical forms. (Continued 5167.6) Variety shown by family
Equivalence relation on Markov chain Lattice of Markov simplifications Markov process / chain equivalence relation on
5164	5165

Summary: Markovian machines simplify on to a lattice.
Summary: Dichotomy versus trichotomy. Dichotomy is best Searching dichotomy is best
5166	5167

Input of mathematical form Transducer mathematical expression as transducer
Summary: Mathematical expressions have inputs and outputs. Summary: Black Box looks at itself and others. The subjective [4]: Can inter-personal relations be translated into Black Box terms? Question asked by Urie Bronfenbrenner, 5169.
5168	5169

Operator "spreading" operator Spreading operator defined

5170	5171

Dependence better Immediate effect algebraic definition

5172	5173

Summary: Algebraic form of "immediate effect". 6070
Integral equation and my methods
5174	5175

History and integral equation Memory and integral equation
Summary: The use and limitations of the integral equation. 5192, 5210, 6036
5176	5177

Summary: Must a person's private "map" be observable to all?
Decay of memory Experience corollary Memory law of experience
5178	5179

Summary: Memory must fail if new information is forced in. 5205
Machine binary relation in Relation binary, in machine
5180	5181



5182	5183

Summary: Machine with input and binary relation.
Isomorphism and prediction Prediction why important
5184	5185

Summary: "A can predict the behaviour of B" is equivalent to "A is isomorphic with B". 5193 Cause Mach on cause Protocol constraint in
Constraint science as constraint, Mach on Laws of nature as constraint Science as constraint Quotations [5]: Scientific law as a re-coding: "Science itself, therefore, may be regarded as a minimal problem, consisting of the completest possible presentment of facts with the least possible expenditive of thought [his italics]." Ernest Mach 5187.9.
5186	5187

Summary: Quotations from Mach. Corespondence as constraint Quotations [57]: "All science has its origin in the needs of life." Ernst Mach 5188.
Summary: The concept of "constraint" includes correspondence, mapping, relation, as special cases. Constraint correspondence as Relation as constraint
5188	5189

Summary: Psychological and anatomical patterns will usually be different. Cf. *Introduction to Cybernetics* S.6/11 Diagram of immediate effects (D.I.E.) anatomical or physiological?
Summary: Two meanings of "input". Input two meanings Protocol general theory of
5190	5191

Constraint in protocal, theory Protocol general theory of
Protocol general theory of
5192	5193



5193+01	5193+02

Canonical equations in protocol Machine in protocol Protocol general theory of
Summary: Constraints in protocols. 5978 Protocol general theory of
5194	5195

Protocol general theory of
Protocol general theory of
5196	5197

Memory in protocol Protocol general theory of
Protocol general theory of
5198	5199

Protocol general theory of
Protocol general theory of
5200	5201

Derivative replacing variable Difference replacing variable Protocol general theory of
Protocol general theory of
5202	5203

Protocol general theory of

5204	5204-A

Protocol general theory of Unobservables replacement by other observations
Protocol general theory of
5205	5206

Protocol general theory of
Protocol general theory of
5207	5208

Summary: Use of observations at times earlier than "immediately preceding." Qualified 5210.3 5228, 5236 Definition and experience Experience and definition Protocol general theory of
Summary: Definition, and the law of experience. 5254 Protocol general theory of
5209	5210

Protocol general theory of
Protocol general theory of
5211	5212

Summary: In the absolute system, the use of earlier states may be inefficient. 5869 Protocol general theory of

5213	5214



5215	5216

Summary: Thoughts provoked by Savage.
Summary: The mapping of the homomorphic machine. Homomorphism [=?µ?^-1] Mapping in homomorphic form
5217	5218

Summary: The structure of the regulator. (Note that the inverse, [w^-1], has appeared at last!) Regulation as inverse Transducer regulator as inverse
Extrapolation Future prediction of Prediction is it possible?
5220	5221


Summary: Must a constraint go on holiday in the future? 5273 Message depends on set
5222	5223

Summary: Example of message that depends on the set it came from. 6275
Step-mechanism ? necessary Ultrastability without step-mechanisms
5224	5225


Summary: Ultrastability in Design and in Introduction reconciled.
5226	5227

Variety real form of
Summary: Real existence of "variety". Metastability meaning of
5228	5229

Summary: Metastability.
Metric Hockett's metric Metric on n-tuples N-tuple metric difference
5230	5231


Input metric in
5232	5233

Summary: Concept of the "distance" between two input states, or two designs. 5506, 6282
Summary: Distance of movement.
5234	5235

Action at a distance Distance action at a distance Noise and action at distance

5236	5237



5238	5239

Summary: Why the scientist should relate the event to what is adjacant rather than distant.
Summary: "Complexity" is a relation between units and construct. Complexity measure of Pre-fabrication in complexity
5240	5241

Entropy zero entropy Homeostasis as noise-correction Noise homeostasis as anti-noise Requisite Variety, Law of as noise-correction
Summary: Law of Requisite Variety as law for suppression of noise. Feedback and persistance Quantum theory and Black Box
5242	5243

1956

Summary: A minute crack at the problem of consciousness. 5275 Conscious mind and feedback Self-awareness and persistance
Summary: The end of the hunt?
5244	5245

Equilibrium homomorphism of Homomorphism of states of equilibrium Order of time
Summary: The machine, when observed only so far as the states of equilibrium it goes to. 5556
5246	5247

Summary: What is a "variable"? Variable nature of
Summary: Mathematical proof "in detail" means going through a sequence of states of equilibrium. Computing machines as noise-destroyer Digital computer as noise-shedder Equilibrium as noise-destroyer Mathematics what is a proof? Proof nature of
5248	5249

Summary: Is there any way by which a person can communicate an infinite amount of information? Consistency in maths
Mathematics consistency in
5250	5251

Summary: "Consistency" in mathematics. Personal notes [24]: My ambition 5252.
Summary: Personal note. Inference nature of
5252	5253

Summary: What is a "logical definition"? Deduction nature of Definition and experience
Infant learning structure Learning without essential variables Structure in world
5254	5255

Summary: The structures of space as habituation. 5289, 5313 Habituation to constraint Space-time structure in Unsolved problems [18]: What can be said of habituation, or of the events in experience, when the set of vectors shows constraint? 5256.
Basin model with lots Habituation model for
5256	5257

Reflex, conditioned machine for

5258	5259

Summary: New machine to demonstrate statistical mechanisms. 5317, 5408, 5526 Imitation mechanistic
Summary: On "imitation". Reproduction more examples
5260	5261

Summary: More "self-reproducing" systems. 5797 Learning in childhood Pain teaches in youth
Summary: Pain educates mostly in childhood. Basin distribution of size
5262	5263


Summary: A contribution to the theory of the distribution of basin size.
5264	5265

Basin distribution of size
Summary: Probability of getting a single basin. 5308 Model Rosenblueth and Wiener
5266	5267

Schizophrenia loss of constraint Psychiatric applications [68]: Is schizophrenia characterised by increasing noisiness? 5268.

5268	5269

Summary: Schizophrenia and noisiness. Games mechanical Pain patient insensitive to Topology concepts of Welch
Mathematics as a model Mathematics product of brain Model theory of mathematical model
5270	5271

Summary: Raiffa et al. describe maths as isomorphic with real systems. 5298, 6260 Canonical equations found by Lotka Delay (in substitution) Lotka on Isomorphism real system and maths

5272	5273


Summary: "Will", and acting now so that certain things shall be. 6321 Conscious mind and feedback
5274	5275

Summary: On consciousness. Pinch as test of identity The subjective [5]: A person may be unconscious of something if he has never conceived of any variation from it 5276. Personal notes [25]: Lazare Courtright gets from "Design for a Brain" a purely emotional message! 5276.
Summary: What good is "Design..."
5276	5277

Summary: Movement between states of equilibrium. Summary: "Yes-no" is a language capable of explaining, or getting, all things, however complex. The Conditioned Reflex [45]: Conditioned reflex probably works by going from equilibrium to equilibrium. 5278.
Summary: Program. DAMS (Dispersive and Multistable System) [96]: DAMS resumed, in Aug '56, after neglect since Mar '53. 5279. Personal notes [27]: Program for the next ten years, as seen in 1956 5279.
5278	5279

Summary: Selection cannot proceed quickly, by dichotomising, in a class that is undefined.
Summary: Example of a case in which information about the initial state gets lost. Absolute system decay of information Information example of decay Initial state decay of information regarding
5280	5281

Summary: Science forswears direct knowledge in favor of the indirect. Epistemology [54]: Primary knowledge is the "raw feel". Science seeks secondary knowledge - the patterned and communicable, which consists only of higher relations among the raw data. 5282. The subjective [6]: The primary data are in the raw feelings. Science works only in higher relations between the raw feelings, for only the higher relations are communicable. 5282.
Summary: "Statistical machine" has two widely different meanings. Machine "random", two types Random two meanings Statistical mechanics two types
5282	5283

Adaptation speeding up adaptation Fringe Habituation as cause of rapid adaptation Speed (of adaptation)

5284	5285

Summary: Fast adaptation can be a manifestation of habituation. 5287.7, 5288, 5342
Summary: How to sample equilibrium states.
5286	5287

Fringe extension of
Summary: The fringe-variables of a system (in this world) are an unbounded set. 5292, 5330 Cortex combinational dynamics of
5288	5289


Summary: Learning structure by habituation to a constraint. Appalling! Habituation is not to a constraint - the ideas have no relation. 5294.4, 5313.3
5290	5291

Summary: How to join motor (and sensory) nerves to a network of part-mechanisms. Now read 3420 and see how much I have advanced in five years!

5292	5293

Summary: Adaptation always consists of one level working with a lower level as units. Play and structure of environment
Summary: In a statistical system, habituation must be used first, in order to provide the permanent entities among which alone trials are worth conducting.
5294	5295

Summary: Why random play should come before serious learning. Axiom the Brick Brick axiom stated
Reducibility getting Higher geometry of fields and matrix theory [28]: Transforming a matrix to reduced form 5297.
5296	5297

Summary: Finding a way of getting a matrix reducible. Theory of nature of
Summary: What is a theory? Cf. Reprint 144 Model nature of
5298	5299

Egyptian steam engine protocol Machine ancient Egyptian Protocol Egyptian Steam Engine Steam engine ancient Egyptian
Summary: Abstract machinery.
5300	5301

Chess machine for
Summary: How to make a brain. Interaction small in multistable systems Retroactive inhibition condition for zero retroactive inhibition
5302	5303

Summary: The multi-stable system must have low interaction on the average. 5312 Summary: On the pains of learning. Dentist Learning necessity The Multistable System [104]: "To accumulate memories that are not affected by intervening events" is equivalent to "There must be small interaction between the events and the memories" 5304.
Constraint and structure Hierarchy (of Bourbaki) and constraint Structure as constraint
5304	5305


Summary: Structure is a form of constraint. 5314, 5496 Summary: To evoke a property, the designer vetoes its negation. 5316, 5331.5 Veto the basic power
5306	5307

Basin distribution of size Machine random machine Mapping random Rubin and Sitgreaves' set of transformations Trajectory distribution of length Transformation random

5308	5309



5310	5311

Memory and internal connexions
Summary: Cross connexions, immediate effects, random transformations, and memory. Experience and learning structure Structure and Law of Experience
5312	5313

Summary: If structure is present in the (set of) inputs, it will turn up in the (set of) states induced by the law of Experience. 5437, 5322 Summary: A reason why a chess-player will automatically tend to learn quick ways of winning. Feedback favours quick adaptation Speed (of adaptation) factor for rapidity The Multistable System [105]: A factor making for fast adaptation 5314.
Pleasure is less effective that pain Success less effective
5314	5315

Summary: Fixing the step-mechanisms on the occurrence of 'success' adds little to the resources of an ultrstable system. 5331.5
Transformation semi-random, many equilibria
5316	5317

Basin model with lots Equilibrium randon transformation with many Field (of substitution) cutting into small basins
Summary: How to chop up a field. 5408, 5524 Personal notes [28]: First advert of "Introduction to Cybernetics", 5319.
5318	5319

Structure learning external structure
Equilibrium and transfer of "structure"
5320	5321


Summary: How does the child get "stucture" from the world around it? It can't keep it out! Copying child and structure
5322	5323

Summary: The "structure" developed by the child (as a result of structure in the world) need not copy the world's structure. Summary: I need not worry further about reversibility. Machine never reversible Reversible process no process is!
Curie's principle
5324	5325

Variety decay of
Summary: Curie's principle is merely a special case of my decay of variety.
5326	5327

Chess analysing ahead DAMS Mark II (Dispersive and Multistable System) [3]: For a machine to "study a move ahead" it has only to be affected by its "provisional" output. 5328.
Summary: "Looking one move ahead does not require special programming.
5328	5329

Fringe
Summary: On the size (duration) of a significant step. 5288 Success less effective
5330	5331

Summary: If punishments only are used, zero punishment is equivalent to reward. Pain is sufficient
Summary: Selection and elimination.
5332	5333

Discrimination in feedback, no general rule Information depends on set

5334	5335

The Multistable System [106]: There is no general rule for how the feedback should be made discrimminating. Only when a particular machine is given can the question proceed to detail. 5336.
Summary: There is no general rule for characterising the more efficient forms of corrective feedback. 5369.2, 5371, 5417, 5539
5336	5337

Transformation moving sequentially DAMS Mark II (Dispersive and Multistable System) [4]: Effect of a transformation being carried out sequentially, (so that e.g. x'₂ depends on x'₁, rather than x₁) 5338.
Summary: Effect of computing a transformation sequentially. Training of full efficiency Trial and error of full efficiency
5338	5339

Summary: Train by situations so matched to the system's present ability that feedbacks of "right" or "wrong" are equally likely. Feedback of full efficiency
Summary: There is no general machine that can be specialised, only a class of individual machines. It is the class that can be broad or narrow. 5507 Machine no general machine, only class
5340	5341

Ultrastability bias to efficiency The Multistable System [107]: The multistable system has a bias towards becoming efficient and quickly adapting. 5342.
Summary: The ultrastable system has an intrinsic bias towards efficiency and quickness in learning. Black box, problem of the poorly observable black box Isomorphism Moore's definition equals mine
5342	5343



5343+01	5343+02

These images are reproduced courtesy of The Estate of W. Ross Ashby. Copyright 1972, 2008 © The Estate of W. Ross Ashby