[ Pobierz całość w formacie PDF ]
.they occur under the heading sampling with replacement.Thus,The second type is well shown by the thermostatically-control-the spin of a coin can give only two results, H or T.A coin spunled water bath, it can be regarded as a regulator, over either shortsix times in succession, however, can give results such as (H, H,or long intervals of time.Over the short interval, the distur-T, H, T, H), or (T, T, H, H, T, H), and so on for 64 possibilities.bance means such an event as the immersion of this flask , and(Compare S.9/9.) its response means what happens over the next minute.ItsWhat is important here is that, in such a set of vectors (in thosebehaviour can be judged good or bad according to what happenedwhose components all come from the same basic class, as in B),248 249AN INTRODUCTION TO CYBERNETICS REGULATING THE VERY LARGE SYSTEMtwo varieties can be distinguished: there is (i) the variety within verified that this sequence, as a protocol, defines the machine withthe basic class (2 for the coin, the number of distinct possible ages input:in B), and (ii) the variety built up by using the basic class n times ± ² ³over (if the vector has n components).In the example of the coin,a ± ± ±the two varieties are 2 and 64.In general, if the variety within theb ² ² ²basic class is k, and the vector has n components, each a memberc ³ ³ ³of the class, then the two varieties are, at most, k, and kn.In par-Thus when the Grand Disturbance is a vector whose componentsticular it should be noticed that if the variety in the basic class hasare all from a basic set of disturbances, the Grand Response cansome limit, then a suitably large value of n will enable the secondeither be a vector of equal variety or the output of a suitablevariety to be made larger than the limit.machine with input.13/9.These considerations are applicable in many cases of regu-13/10.Suppose that the regulation discussed throughout Part III islation.Suppose, for definiteness, that the water bath may bethe responsibility of some entity Q, often the possessor of theaffected in each minute by one of the three individual distur-essential variables E.Through the previous chapters we have stud-bances:ied how the regulator R must behave.We have now seen that in thecase when the disturbances are repetitive, Q has the option of(a) a draught of air cooling it,either being the regulator (i.e.acting as R) or of building a machine(b) sunshine warming it,that, once built, will act as R and will carry out a regulation of(c) a cold object being immersed in it.indefinite length without further action by Q.We have thus arrivedThe variety is three, but this number is hardly representative ofat the question: should Q achieve the regulation directly, by histhe variety that will actually occur over a long time.Over a year,own actions, or should he build a machine to undertake the work?say, the Grand Disturbance is a long vector, with perhaps someThe question would also have arisen for another reason.Fromhundreds of components.Thus one Grand Disturbance might bethe beginning of Part III we took for granted that the regulatorthe vector (i.e.the sequence) with 400 components:existed, and we then asked what properties it must have.Nothingwas said about how the regulator came to be made, about the fac-(a, b, a, b, b, a, c, b, b, c, c, b, b, & c, b, a, b).tors that brought it into existence.Thus, having seen in S.10/5 howAnd if the individually correct responses are, respectively ±, ²advantageous it would be if the organism could have a regulator,and ³ , then the Grand Response appropriate to this particular Dis-we showed no means by which the advantage could be gained.turbance would be the vector (i.e.sequence)For both these reasons we must now start to consider how a reg-ulatory machine is actually to be designed and made.Here we(±, ², ±, ², ², ±, ³, ², ², ³, ³, ², ², & ³, ², ±, ² ).shall be thinking not so much of the engineer at his bench as of theIf there is no constraint in the Disturbance from component tobrain that, if it is to achieve regulation in its learned reactions,component as one goes from left to right, the whole set of possiblemust somehow cause the development of regulatory machineryDisturbances has variety of 3400; and the Grand Response mustwithin the nervous material available; or of the sociologist whohave at least as much if full regulation is to be obtained.wants a regulatory organisation to bring harmony into society.We now come to the point: the double sequence, as it occurredTo understand what is involved, we must look more closely atin time, shows the characteristic constraint of a machine, i.e.itwhat is implied, in principle, in the designing of a regulatorydefines a machine up to an isomorphism.Thus, in the examplemachine.just given, the events occurred in the order, from left to right:DESIGNING THE REGULATORa b a b b a c b b c c & , etc.3|11.Design as communication.Let us forget, temporarily, all± ² ± ² ² ± ³ ² ² ³ ³ & , etc.about regulation , and turn simply to certain questions related to(though not necessarily at equal time-intervals).It is now easilythe design and construction of a machine, any machine.250 251AN INTRODUCTION TO CYBERNETICS REGULATING THE VERY LARGE SYSTEMOur treatment of it, while losing nothing in precision, must be It will be noticed that the concepts of design or construction arevery broad that is to say, abstract for, as biologists, we want to essentially applicable to sets, in spite of the common linguistic useconsider machines of far wider type than those of steel and brass.of the singular.(Compare S.7/3.) Thus the gene-pattern deter-Within the formula mines the form of the heart is a shorthand way of saying that ele-ments in the set of gene-patterns among different species can beEntity &! designs machine Mput into correspondence with those in the set of possible hearts inwe want to include such cases as the various species, like the wires at the two ends of a telephonecable.Thus the act of designing or making a machine is(l) The genes determining the formation of the heart.essentially an act of communication from Maker to Made, and the(2) A mechanic making a bicycle.principles of communication theory apply to it.In particular the(3) One part of the brain determining the internal connexions inmeasures that were developed for treating the case in which vari-a nerve-net.ous possible messages are reduced to one message can now be(4) A works-manager laying out a factory to get production go-applied to the case when various possible machines are reduced toing along certain lines.one machine.(5) A mathematician programming an automatic computer toA useful conceptual device for forcing this aspect into promi-behave in a certain way.nence is to imagine that the act of designing has to take placethrough the telephone, or by some other specific channel.TheWhat we shall be concerned with, if we hold to the cyberneticquantities of variety can then readily be identified by identifica-point of view, is not the more obvious processes of shaping ortion of the actual quantity of information that will have to be trans-assembling pieces of matter, but with the less obvious questions ofmitted [ Pobierz caÅ‚ość w formacie PDF ]