The Sciences of the Artificial

tags

Herbert Simon

Notes

Preface to Second Edition

NOTER_PAGE: (9 . 0.12135922330097088)

The thesis is that certain phenomena are "artificial" in a very specific sense: they are as they are only because of a system's being moulded, by goals or purposes, to the environment in which it lives.

NOTER_PAGE: (9 . 0.45469255663430425)

artificial phenomena have an air of "contingency"

NOTER_PAGE: (9 . 0.5283171521035599)

how empirical propositions can be made at all about systems that, given different circumstances, might be quite other than they are.

NOTER_PAGE: (9 . 0.6820388349514563)

1 Understanding the Natural and the Artificial Worlds

NOTER_PAGE: (13 . 0.12216828478964402)

The central task of a natural science is to make the wonderful commonplace: to show that complexity, correctly viewed, is only a mask for simplicity; to find pattern hidden in apparent chaos.

NOTER_PAGE: (13 . 0.3122977346278317)

Wonder, en is gheen wonder

NOTER_PAGE: (13 . 0.6731391585760518)

This is the task of natural science: to show that the wonderful is not incomprehensible

NOTER_PAGE: (13 . 0.7402912621359223)

the significant part of the environment consists mostly of strings of artifacts called "symbols"

NOTER_PAGE: (14 . 0.691747572815534)

They are what they are in order to satisfy our desire

NOTER_PAGE: (15 . 0.5469255663430421)

objects and phenomena in which human purpose as well as natural law are embodied

NOTER_PAGE: (15 . 0.6326860841423948)

using "artificial" in as neutral a sense as possible, as meaning man-made as opposed to natural.

NOTER_PAGE: (16 . 0.255663430420712)

The engineer, and more generally the designer, is concerned with how things ought to be how they ought to be in order to attain goals and to function

NOTER_PAGE: (16 . 0.6076051779935275)

dichotomy between normative and descriptive

NOTER_PAGE: (17 . 0.16990291262135923)

four indicia that distinguish the artificial

NOTER_PAGE: (17 . 0.30258899676375406)

a science of the artificial that would depend on the relative simplicity of the interface as its primary source of abstraction and generality.

NOTER_PAGE: (21 . 0.24757281553398058)

Often we shall have to be satisfied with meeting the design objectives only approximately. Then the properties of the inner system will "show through."

NOTER_PAGE: (24 . 0.4563106796116505)

digital computer has greatly extended the range of systems whose behavior can be imitated.

NOTER_PAGE: (25 . 0.5040453074433657)

the crucial question about simulation: How can a simulation ever tell us anything that we do not already know?

NOTER_PAGE: (26 . 0.5566343042071198)

All correct reasoning is a grand system of tautologies, but only God can make direct use of that fact.

NOTER_PAGE: (27 . 0.15857605177993528)

a digital computer. It is truly protean, for almost the only ones of its properties that are detectable in its behavior (when it is operating properly!) are the organizational properties.

NOTER_PAGE: (29 . 0.6262135922330098)

Computers As Empirical Objects

NOTER_PAGE: (31 . 0.284789644012945)

The computer is a member of an important family of artifacts called symbol systems

NOTER_PAGE: (33 . 0.6067961165048544)

A physical symbol system is a machine that, as it moves through time, produces an evolving collection of symbol structures.

NOTER_PAGE: (34 . 0.377831715210356)

Symbol systems are called "physical" to remind the reader that they exist as real- world devices, fabricated of glass and metal (computers) or flesh and blood (brains).

NOTER_PAGE: (34 . 0.7799352750809062)

2 Economic Rationality: Adaptive Artifice

NOTER_PAGE: (37 . 0.1156957928802589)

It is widely taught in both senses in business schools and universities, just as if it described what goes on, or could go on, in the real world. Alas, the picture is far too simple to fit reality.

NOTER_PAGE: (38 . 0.3244336569579288)

human choices are not consistent and transitive, as they would be if a utility function existed

NOTER_PAGE: (41 . 0.3956310679611651)

Markets are only one, however, among the spectrum of mechanisms of coordination on which any society relies.

NOTER_PAGE: (43 . 0.14320388349514565)

The economic units in capitalist societies are mostly business firms, which are themselves hierarchic organizations, some of enormous size, that make almost negligible use of markets in their internal functioning.

NOTER_PAGE: (43 . 0.6407766990291263)

The optimality theorems stretch credibility, so far as real-world markets are concerned, because they require substantive rationality of the kinds we found implausible in our examination of the theory of the firm.

NOTER_PAGE: (45 . 0.2127831715210356)

Samuel Johnson said of the dancing dog, "The marvel is not that it dances well, but that it dances at all"

NOTER_PAGE: (45 . 0.40291262135922334)

the marvel is not that markets optimize (they don't) but that they often clear.

NOTER_PAGE: (45 . 0.43284789644012944)

market processes commend themselves primarily because they avoid placing on a central planning mechanism a burden of calculation that such a mechanism, however well buttressed by the largest computers, could not sustain.

NOTER_PAGE: (46 . 0.5266990291262136)

how little the individual participants need to know in order to be able to take the right action.

NOTER_PAGE: (46 . 0.8705501618122978)

Although the presence of uncertainty does not make intelligent choice impossible, it places a premium on robust adaptive procedures instead of optimizing strategies

NOTER_PAGE: (47 . 0.709546925566343)

Feed forward in markets can become especially destabilizing when each actor tries to anticipate the actions of the others (and hence their expectations).

NOTER_PAGE: (48 . 0.220873786407767)

in actual experiments with the game, it turns out that cooperative behavior occurs quite frequently

NOTER_PAGE: (50 . 0.13430420711974111)

Market institutions are workable (but not optimal) well beyond that range of situations precisely because the limits on human abilities to compute possible scenarios of complex interaction prevent an infinite regress of mutual outguessing.

NOTER_PAGE: (50 . 0.4911003236245955)

rationality is effectively undefinable when competitive actors have unlimited computational capabilities […] the problem does not arise as acutely in a world, like the real world, of bounded rationality.

NOTER_PAGE: (50 . 0.5833333333333334)

Uncertainty calls for flexibility, but markets do not always provide the greatest flexibility in the face of uncertainty.

NOTER_PAGE: (55 . 0.07766990291262137)

Docile people do not have to learn about hot stoves by touching them.

NOTER_PAGE: (57 . 0.18290441176470587)

The evolution and future of such systems can only be understood from a knowledge of their histories.

NOTER_PAGE: (59 . 0.8676470588235294)

3 The Psychology of Thinking: Embedding Artifice in Nature

NOTER_PAGE: (63 . 0.11029411764705882)

Viewed as a geometric figure, the ant's path is irregular, complex, hard to describe. But its complexity is really a complexity in the surface of the beach, not a complexity in the ant.

NOTER_PAGE: (63 . 0.6920955882352942)

Human beings, viewed as behaving systems, are quite simple. The apparent complexity of our behavior over time is largely a reflection of the complexity of the environment in which we find ourselves.

NOTER_PAGE: (65 . 0.20220588235294118)

proposed scheme requires finding the contradictions implied by an assignment. This means of course the "relatively direct" contradictions, for if we had a rapid process capable of detecting all inconsistent implications, direct or indirect, it would find the problem solution almost at once.

NOTER_PAGE: (68 . 0.7647058823529411)

human beings do not always discover for themselves clever strategies that they could readily be taught

NOTER_PAGE: (73 . 0.2389705882352941)

4 Remembering and Learning: Memory As Environment for Thought

NOTER_PAGE: (97 . 0.1213235294117647)

Intuition is a genuine enough phenomenon which can be explained rather simply: most intuitive leaps are acts of recognition.

NOTER_PAGE: (101 . 0.20220588235294118)

From what is known about the rates at which people can store new information in long-term memory, 50,000 chunks is not an unconscionable amount of knowledge to acquire in a decade, say, of professional training.

NOTER_PAGE: (103 . 0.3428308823529412)

the most talented people require approximately a decade to reach top professional proficiency

NOTER_PAGE: (103 . 0.7003676470588235)

There is a constant competition between the elaboration of knowledge and its compression into more parsimonious form by theories.

NOTER_PAGE: (104 . 0.7132352941176471)

it is not safe to say that the professional chemist must learn more today than a half century ago, before the general laws of quantum mechanics were announced.

NOTER_PAGE: (104 0.7621359223300971 . 0.13877118644067796)

Production Systems

NOTER_PAGE: (114 . 0.0900735294117647)

The test that something has been discovered is that something new has emerged that could not have been predicted

NOTER_PAGE: (118 . 0.19852941176470587)

Another discovery system of considerable interest is the BACON program, in its several forms, which is capable of discovering invariants in bodies of numerical data.

NOTER_PAGE: (119 . 0.12224264705882353)

adding artificial intelligence to the tools of science.

NOTER_PAGE: (120 . 0.1488970588235294)

Every problem-solving effort must begin with creating a representation for the problem

NOTER_PAGE: (120 . 0.5349264705882353)

Focus of attention is the key to success

NOTER_PAGE: (121 . 0.5790441176470588)

5 The Science of Design: Creating the Artificial

NOTER_PAGE: (123 . 0.12316176470588235)

Everyone designs who devises courses of action aimed at changing existing situations into preferred ones.

NOTER_PAGE: (123 . 0.2895220588235294)

Design, so construed, is the core of all professional training; it is the principal mark that distinguishes the professions from the sciences.

NOTER_PAGE: (123 . 0.3841911764705882)

As professional schools, including the independent engineering schools, were more and more absorbed into the general culture of the university, they hankered after academic respectability

NOTER_PAGE: (124 . 0.38786764705882354)

Since the concern of standard logic is with declarative statements, it is well suited for assertions about the world and for inferences from those assertions. Design, on the other hand, is concerned with how things ought to be, with devising artifacts to attain goals.

NOTER_PAGE: (126 . 0.5983455882352942)

the requirements of design can be met fully by a modest adaptation of ordinary declarative logic. Thus a special logic of imperatives is unnecessary.

NOTER_PAGE: (127 . 0.6397058823529411)

Solving a problem simply means representing it so as to make the solution transparent.

NOTER_PAGE: (144 . 0.4911003236245955)

people or at least their intellective component may be relatively simple, that most of the complexity of their behavior may be drawn from their environment

NOTER_PAGE: (150 . 0.24757281553398058)

6 Social Planning: Designing the Evolving Artifact

NOTER_PAGE: (151 . 0.11607910576096303)

Going to the Moon was a complex matter along only one dimension: it challenged our technological capabilities.

NOTER_PAGE: (151 . 0.653150343106675)

condition of their success was that they were evaluated against limited objectives

NOTER_PAGE: (152 . 0.6562694946974423)

At least six different, and largely contradictory, conceptions were offered for the agency by the persons who were initially recruited to organize and manage it.

NOTER_PAGE: (154 . 0.13349968808484092)

What was needed was not so much a "correct" conceptualization as one that could be understood by all the participants and that would facilitate action rather than paralyze it.

NOTER_PAGE: (155 . 0.25015595757953835)

A design representation suitable to a world in which the scarce factor is information may be exactly the wrong one for a world in which the scarce factor is attention.

NOTER_PAGE: (156 . 0.2195882719900187)

there was no way in which the hypothetical cost-benefit analysis scheme could be applied literally. Nevertheless the scheme provided a conceptual framework

NOTER_PAGE: (158 . 0.08796007485963818)

One may regard "defensibility" as a weak standard for a decision on a matter as consequential as automobile emissions. But it is probably the strictest standard we can generally satisfy with real-world problems of this complexity.

NOTER_PAGE: (158 . 0.33499688084840923)

Numbers are not the name of this game but rather representational structures that permit functional reasoning, however qualitative it may be.

NOTER_PAGE: (158 . 0.47723019338739864)

predictions are commonly the weakest points in our armor of fact.

NOTER_PAGE: (159 . 0.21397379912663755)

require either a theoretical understanding of the phenomena to be predicted, as a basis for the prediction model, or phenomena that are sufficiently regular that they can simply be extrapolated.

NOTER_PAGE: (159 . 0.2601372426699938)

reliable data about the initial conditions

NOTER_PAGE: (159 . 0.3917654398003743)

Club of Rome report, which predicted a twenty- first century doomsday of overpopulation, resource exhaustion, and famine.

NOTER_PAGE: (159 . 0.6743605739238927)

We do not want to know when disaster is going to strike but how to avoid it.

NOTER_PAGE: (159 . 0.8128509045539614)

Predicting the exact course of global warming is a thankless task. Much more feasible and useful is generating alternative policies

NOTER_PAGE: (160 . 0.3587024329382408)

Future contingencies that have no implications for present commitment have no relevance to design.

NOTER_PAGE: (160 . 0.8303181534622582)

active, feed forward control, using predictions, can throw a system into undamped oscillation

NOTER_PAGE: (161 . 0.7205240174672489)

Because of the possible destabilizing effects of taking inaccurate predictive data too seriously, it is sometimes advantageous to omit prediction entirely

NOTER_PAGE: (161 . 0.7704304429195259)

"Who is the client?" when speaking of the design of large social systems.

NOTER_PAGE: (162 . 0.12663755458515283)

Developments in technology give professionals the power to produce larger and broader effects at the same time that they become more clearly aware of the remote consequences of their prescriptions.

NOTER_PAGE: (162 . 0.5864004990642545)

professionals find they no longer serve individual clients but are employed directly by agencies of the state.

NOTER_PAGE: (163 . 0.11291328758577666)

In city planning, for example, the boundary between the design of physical structures and the design of social systems dissolves almost completely.

NOTER_PAGE: (163 . 0.825951341235184)

relatively rare for social planning and policy discussions to include in any systematic way the possible "gaming" responses to plans.

NOTER_PAGE: (166 . 0.33624454148471616)

If we were to assign a single cause to our good fortune, we would have to attribute it to being born in the right place at the right time

NOTER_PAGE: (167 . 0.13583617747440274)

Each of us sits in a long dark hall within a circle of light cast by a small lamp. The lamplight penetrates a few feet up and down the hall, then rapidly attenuates, diluted by the vast darkness of future and past that surrounds it.

NOTER_PAGE: (168 . 0.2143344709897611)

History, archaeology, geology, and astronomy provide us with narrow beams that penetrate immense distances down the hallway of the past but illuminate it only fit fully a statesman or philosopher here, a battle there, some hominoid bones buried with pieces of chipped stone, fossils embedded in ancient rock, rumors of a great explosion.

NOTER_PAGE: (168 . 0.49351535836177474)

If our decisions depended equally upon their remote and their proximate consequences, we could never act but would be forever lost in thought. By applying a heavy discount factor to events, attenuating them with their remoteness in time and space, we reduce our problems of choice to a size commensurate with our limited computing capabilities.

NOTER_PAGE: (169 . 0.28259385665529013)

bias our investments in the direction of structures that can be shifted from one use to another, and to knowledge that is fundamental enough not soon to be outmoded

NOTER_PAGE: (170 . 0.48668941979522184)

modern analytic tools have taught us how to detect good and evil in minute amounts and at immense distances in time and space.

NOTER_PAGE: (171 . 0.5972696245733788)

there has been a genuine shift in our orientation to time and a significant lengthening in time perspectives.

NOTER_PAGE: (172 . 0.1597269624573379)

if we live in a time that is sometimes pessimistic about technology, it is because we have learned to look farther than our arms can reach.

NOTER_PAGE: (172 . 0.35426621160409555)

From a pragmatic standpoint we are concerned with the future because securing a satisfactory future may require actions in the present. Any interest in the future that goes beyond this call for present action has to be charged to pure curiosity.

NOTER_PAGE: (173 . 0.20955631399317406)

We saw there that search guided by only the most general heuristics of "interestingness" or novelty is a fully realizable activity.

NOTER_PAGE: (174 . 0.5556313993174061)

The painting process is a process of cyclical interaction between painter and canvas in which current goals lead to new applications of paint, while the gradually changing pattern suggests new goals.

NOTER_PAGE: (175 . 0.48464163822525597)

If we conceive human beings as having some kind of alterable capacity for enjoyment and appreciation of life, then surely it is a reasonable goal for social decision to invest in that capacity for future enjoyment.

NOTER_PAGE: (176 . 0.33515358361774744)

one goal of planning may be the design activity itself. The act of envisioning possibilities and elaborating them is itself a pleasurable and valuable experience.

NOTER_PAGE: (176 . 0.4839590443686007)

7 Alternative Views of Complexity

NOTER_PAGE: (180 . 0.12150170648464163)

The period during and just after World War II saw the emergence of what Norbert Wiener dubbed "cybernetics"

NOTER_PAGE: (183 . 0.4928327645051195)

Holism was brought into confrontation with reductionism in a way that had never been possible before

NOTER_PAGE: (184 . 0.2627986348122867)

During these postwar years, a number of proposals were advanced for the development of "general systems theory," that, abstracting from the special properties of physical, biological, or social systems, would apply to all of them. We might well feel that, while the goal is laudable, systems of these diverse kinds could hardly be expected to have any nontrivial properties in common.

NOTER_PAGE: (184 . 0.3733788395904437)

8 The Architecture of Complexity: Hierarchic Systems

NOTER_PAGE: (194 . 0.11945392491467577)

the whole is more than the sum of the parts in the weak but important pragmatic sense that, given the properties of the parts and the laws of their interaction, it is not a trivial matter to infer the properties of the whole.

NOTER_PAGE: (195 . 0.11262798634812286)

Physics makes much use of the concept of "elementary particle," although particles have a disconcerting tendency not to remain elementary very long.

NOTER_PAGE: (196 . 0.08395904436860069)

The time required for the evolution of a complex form from simple elements depends critically on the numbers and distribution of potential intermediate stable forms.

NOTER_PAGE: (201 . 0.4709897610921502)

Among possible complex forms, hierarchies are the ones that have the time to evolve.

NOTER_PAGE: (208 . 0.13651877133105803)

Nearly Decomposable Systems

NOTER_PAGE: (208 . 0.31467576791808877)

As a second approximation we may move to a theory of nearly decomposable systems, in which the interactions among the subsystems are weak but not negligible.

NOTER_PAGE: (208 . 0.7440273037542662)

(1) in a nearly decomposable system the short-run behavior of each of the component subsystems is approximately independent of the short-run behavior of the other components

NOTER_PAGE: (209 . 0.08873720136518772)

(2) in the long run the behavior of any one of the components depends in only an aggregate way on the behavior of the other components.

NOTER_PAGE: (209 . 0.157679180887372)

whether we are able to understand the world because it is hierarchic or whether it appears hierarchic because those aspects of it which are not elude our understanding and observation.

NOTER_PAGE: (219 . 0.08600682593856655)

Plato, in the Meno, argued that all learning is remembering. He could not otherwise explain how we can discover or recognize the answer to a problem unless we already know the answer.

NOTER_PAGE: (221 . 0.8402730375426621)

we might expect ontogeny partially to recapitulate phylogeny in evolving systems whose descriptions are stored in a process language

NOTER_PAGE: (226 . 0.15085324232081912)

progress from elementary to advanced courses is to a considerable extent a progress through the conceptual history of the science itself.

NOTER_PAGE: (226 . 0.26621160409556316)

We do not teach the phlogiston theory in chemistry in order later to correct it.

NOTER_PAGE: (226 . 0.30921501706484644)

hierarchies have a property, near decomposability, that greatly simplifies their behavior.

NOTER_PAGE: (227 . 0.2935153583617747)