[paper] darwinian design thememetic evolution

7/27/2019 [Paper] Darwinian Design TheMemetic Evolution

1/17


2/17

Darwinian esign:TheMemeticEvolutionofDesign deasJohnZ.Langrish

Aversionfthispaper as presentedoaHistoryndTheoryf Design onferencesponsoredy he Neohellenic esearchInstitutend he DesignResearchociety(Syros, uly 003). tsoriginalitlewas"TheEvolutionf Design deas-Memetic otProgressive."

IntroductionThere seems to have been a recent slight increase in the numberof design papers with the word, "evolution" in their titles.Unfortunately, these papers are either vague about what is meantby this word, or they use the word in a non-Darwinian sense whichowes more to Spencer's version of progressive evolution than to theprocess of natural selection.

One interesting example is a paper by A. Can Ozcan, whowrites:

Let's assume that the one we know as Darwin is born in ourtimes and he is very curious not about species but designedobjects and artifacts. Instead of looking at birds he is look-ing at refrigerators, cars, kettles, microphones, bicycles. Ournumber one question is whether he would come up withsimilar principles of evolution like selection of the fittestor progression from simplicity to complexity for designedobjects.'

My short answer to that question is an emphatic "No." The longeranswer is that Charles Darwin did not come up with "principles ofevolution," and if he had done so then progress towards complexitywould not be one of them. The original full title of his great workwas On the Origin of Species by Means of Natural Selection-nothingabout "evolution." In fact, the word "evolution" is only used oncein the first edition. He originally intended to call this work just"Natural Selection," and a Darwinian theory is one based on naturalselection-not on some inevitable force for progress.

The term "survival of the fittest" was used by HerbertSpencer before Darwin was persuaded to copy it in later editionsof his work. The notion of progress from simple to complex is a keypart of Spencer's evolution, but it does not correspond with whatwe know. This paper suggests that Spencerian notions of progres-sive evolution have dominated discussions of evolution in design,and now it is time to examine what a Darwinian theory of designevolution might look like. Darwin, of course, did not know anythingabout genes, genetics, or mutation. The term neo-Darwinism is used

1 A.CanOzcan, nEvolutionaryApproachorDesign-ContradictoryorComplementaryithHistory(3rdInternationalonferencenDesignHistoryndDesign tudies,stanbul,2002.)

( 2004 John Z.Langrish4 Design Issues: Volume 0, Number Autumn 004


3/17

to mean Darwin's natural selection plus genes. It is not suggestedthat design is somehow genetic. Design evolution is the evolutionof ideas, and the Darwinian evolution of ideas is called "memetics"from the concept of self-replicating ideas called memes by RichardDawkins.2

Four Arguments Against Evolutionary DesignA good example of the way in which design historians equate evolu-tion with Spencerian notions of progress is provided by AdrianForty's Objectsof Desire.3

In his otherwise excellent attempt to tackle the problem ofwhy artifacts are the way they are, Forty dismisses evolutionaryexplanations of change on the following grounds:

Historians of design have often tried to get around theproblem [of explanations involving creative individuals]by attributing the changes to some sort of evolutionaryprocess, as if manufactured goods were plants or animals.Changes in design are described as if they were mutationsin the development of products, stages in a progressiveevolution towards their most perfect form. But artifacts donot have a life of their own, and there is no evidence for alaw of natural or mechanical selection to propel them in thedirection of progress. The design of manufactured goodsis determined not by some internal genetic structure butby the people and the industries that make them and therelationships of these people and industries to the society inwhich the products are to be sold.3

Forty has four arguments against what he calls evolution. They maybe good arguments against vague ideas of Spencerian evolution, butthey are not valid arguments against Darwinian change. His fourarguments are:

1. The progress argument. This has nothing to do withDarwinian change, but Forty does not restricthimself to the progresstowards complexity mistake: he adds the astonishing "aprogressiveevolution towards their most perfect form." There is no such thingas a perfect mammal, perfect kettle, perfect car,or perfect tree. In allcases, they exist as different varieties which have to fit into differentenvironments. "Progressive evolution towards a perfect form" is anexample of what ErnstMayr refers to as "finalism"or "thebelief thatthe living world has the propensity to move towards ever greaterperfection." According to Mayr, supporters of finalism "postulatedthe existence of some built in force... but Darwin emphaticallyrejected such obscure forces."

2. 'Artifacts do not have a life of their own." This argumentis also known as the "machines don't mate" argument. The shortanswer to this is that the evolution of design ideas is the issue, andthat ideas do have a "life."

2 RichardDawkins,"Memes:TheNewReplicators," hapter 1 in TheSelfishGene Oxford: xfordUniversity ress,1976), 189-201.

3 AdrianForty,Objectsof Desire:Designand Society Since 1750(London:hamesandHudson,1986),8.

4 ErnstMayr,WhatEvolution s(London:BasicBooks,edition:Phoenix, 001),82.

Designssues:Volume0,Number Autumn004 5


4/17

The machines don't mate argument was answered convinc-ingly by Samuel Butler more than a hundred years ago. Butler'sarguments were written in three chapters of his novel Erewhon(anagram of nowhere, and pronounced with three syllables). Oneof Butler's responses to "machines don't mate" was "Does anyonesay that the red clover has no reproductive system because thehumble bee must aid and abet it before it can reproduce? No one."Machines use humans to "aid and abet" them. He makes the obvi-ous points about individual machines requiring feeding and tendingby humans, but he also makes the much more subtle point that theimprovement of machinery relies on competition, the destruction ofinferior machines, and the creation of better machines. These threetasks all require the enslavement of humans. In Butler's words:

The lower animals progress because they struggle withone another; the weaker die, the stronger breed and trans-mit their strength. The machines themselves being unableto struggle have got man to do their struggling for them;as long as he fulfils this function duly, all goes well withhim-at least he thinks so; but the moment he fails to do hisbest for the advancement of machinery by encouraging thegood and destroying the bad, he is left behind in the race ofcompetition; and this means that he will be made uncom-fortable in a variety of ways and perhaps die.5

Butler, of course, did believe in progress, and this lead him to partcompany with Darwin.

A modern answer to the "machines don't mate" argumentwould involve the fact that life on earth went on for about one-thou-sand million years before sexual reproduction appeared. The earlybacteria mixed up their genetic material in a variety of ways includ-ing lateral transfer. This is more like the way in which design ideasmix together. Some of the bacteria which are around today are verysimilar to the ancient bacteria that first appeared about thirty-eighthundred million years ago. This does not suggest that Spencer wasright in his ideas of progression.

3. The law of propulsion argument. Natural selection isnot a law like the law of gravity: it does not propel things in somepredetermined direction. It is a filter, which is different. If we have amixture of different sizes and shapes of things being shaken on top ofa sieve, then some things will pass through the sieve and some willnot. The force at work here is the force of gravity, which is impartial.The sieve which "selects"things as being below a certain size is not a"force."Some things just pass through it, and some things don't. Andthere is a little luck involved here in that some small things whichought to pass through the sieve don't because they get stuck, andsome large things which happen to be long and very thin manage towriggle through. So there is no precise prediction of the separation.SamuelButler,Erewhon: rOver he

Range London:onathanCape,1872).

6 Designssues:Volume0, Number Autumn004


5/17

It took time forthe differencebetween naturalselection andsome kind of propulsiveforceto be appreciated. n the early yearsof the twentieth century,there were several people postulating a"force."One was the Frenchphilosopher Henri Bergson,whose"1creativevolution"was propelledby anelanvital-a vitalimpetus.According o the FrencharthistorianGermainBazin,Bergson nflu-enced arthistory.Referring o this influenceof the "philosophyofBergson,"Bazin claims:

Arthistorians, ollowing a certain inalisttendencywhichshowed itselfparticularlyn neo-vitalistdoctrines,begantoseek the determining actors n the work of artno longerincircumstances utside the work,but in the artisticactivityitself. They credited his activitywith a capacity ordevel-opment orexpansionof its own, to be understood ike lifein terms of a "creative volution"workingtowardsa moreefficientuse of its inherentproperties.6

Bergson'sevolution, like Spencer's,has no evidence in its favor.Natural selectionhas morethan a hundredyearsof evidence in itssupport.

4.The argument hatmanufactured oods do not have some"internalgenetic structure."Theshortanswerto this is to point outthatDarwinknew nothingat all aboutgenes or genetics,so what-ever is meantby Darwinianchangedoes not have to include some"internal eneticstructure." arwin,of course,was aware hatsome-thinghad to be passedon fromone generation o another; therwisenaturalselection would not work. However, his ideas about thenature of this "something"were confused.Themoderntermfor a"~something"hatgets passedon is a "replicator"s popularizedbyRichardDawkins, who points out thatthere must have been chemi-cal replicatorsbeforethe emergenceof DNA, and also thathumansocietyrestson a new typeof replicatorwhich he calls a "meme"-areplicating dea.2 deas thatget copied,modified, and stucktogetherwith other deas can formthe basis of a Darwinian heory of chang-ing design.Thestudy of replicating deas is called"memetics," utbeforemoving to a discussion of memetics, it is necessary to saymore about the ideas of SpencerandLamarck.SpencerianProgressThe idea that thereis some propulsive force or "law"of progressbehind evolution seems to have arrivedin the history of art anddesignas a resultof thewritingsof HerbertSpencer. n a 1961paper,ThomasMunro claimed thatSpencerproduced "thefirst detailedsystematicattemptto fit the historyof art into a naturalisticheoryof evolution."7Increasing complexity,accordingto Spencer,was achangefrom the homogeneousto the heterogeneous,and from theindefiniteto the definite. Thedevelopment of the arts,he believed,

6 Germainazin,A oncise istoryf Art(London:hamesndHudson,958), 22

7 ThomasMunro,DoheArtsEvolve?SomeRecentConflictingnswers,"JoumalfAesthetics ndArtCriticism(Summer961):07-417.

Design Issues: Volume 0, Number Autumn 004 7


6/17

illustrated his tendency and thus exemplifiedthe largerprocessofmental and socialevolution.

Spencerwrote about this in 1857, wo yearsbeforeDarwin'sOrigin.Thetitleof his 1857essaywas "Progress,ts Law andCause."He called the process of increasingcomplexity"progress,"whichthen became "evolution."Theevolutionary processtook in every-thing fromthe starsto the arts.Lateron, theprocessof social evolu-tion came to be describedas "cultural volution."

After discussing notions of evolution in the history of art,Munroasks:

... if the term "evolution n art" s so ambiguous,so loadedwith inconsistentmeanings, s itusable at all in scholarlydiscussion?Would t be betterto findanother erm,or a setof them?"

My answer is "yes"; t would be much betterif we stopped usingevolution and used "Darwinianchange"to signify descent withmodification under the influence of natural selection. If someother kind of process is under discussion, then other terms exist.Spencerianchangecould be used for an inevitableprocessleadingto greatercomplexityandimprovement.Lamarckianhangecouldsignifya processwhereby changeresultsfromstrivingforimprove-ment, and the further ransmissionof suchimprovement.

Makinga clear distinction between Spencerian notions ofchangeand Darwinianchangeis essentialif evolutionaryaccountsof design changeareto be treatedseriously.Thesedays,mosthisto-rians rejecthistorical "forces"as a meaningful concept,and manyareunhappywithnotions of progress.Historicism forcesof history)and Whig history (things get better over time) both have beendiscarded.SinceSpencer's deas includeboth "forces"and "thingsgetting better,"t is not surprising hatSpenceriannotionsof evolu-tion have beenrejected.However,Darwinianevolutiondependsonnaturalselection,which is a filternot a "force"and does not claimthat change must be progressive (though it might seem to be onoccasion).As long as historiansconfuse evolution with Spencerianchange,evolution is going to be rejected,as it has been by AdrianForty.The Darwinianalternativehas not beengiven a chance.

Another reason for rejectingSpencerian notions of prog-ress throughincreasing complexity is that they just do not fit thefacts.ManywritersbeforeDarwin,includinghis own grandfather,ErasmusDarwin, had notions of a progressive gradual change.Lamarckwas so keen on progressivecomplexity thatwhen it waspointed out that there were some new, simplerorganisms,he wasforced to suggest that they must be the product of spontaneousgeneration.

Darwin's natural selection is different; t is notessentiallyprogressive: it is more in accord with what we observe in naturewhere there aremany examples of things becoming less complex.

8 Design ssues:Volume0, NumberAutumn004


7/17

The remoteancestorsof thehorse,forexample,had five toes whichbecame three and then one. At some stages in the complexhistoryof horse species, there were species which became smallerwhileothersbecame arger.Therewas no simplelineardevelopment roma smalldog-likecreature o the modern horse.Ithas takentimeforthis fact to be appreciated y non-biologists.Thefaultforthis lies inmuseumexhibitssuch as those thatused the evolution of the horseas a visualillustration f evolution,givingtheimpressionof progressin a particulardirection.

In 1959,the NaturalHistory Museumin Londonwas illus-tratingevolutionwith, amongotherthings, the horseprogression.Thisexhibit had four skeletons, the Hyracotheriumor Eohippus,which was aboutthe height of a fox terrier, he Miohippus,whichwas about wo feethigh (shoulderheight), he Pliocenehorsewhich,was aboutfour feethigh, and the modernEquus.Thefeet changedgraduallyfromhaving toes in the Eohippusto havinghoofs in theEquus.Similarly, he teethseemed to changegradually.

A slight hintthatthingswerenot quite as simpleassuggestedby the gradualprogressionof the skeletonswas given in thewordsof abookletaccompanying he exhibitwhichclaimed thattheearlyskeletonscamefromEurope,but the "genusEquusfirstappearedaboutamillionyearsagoin NorthAmerica,whence t spreadrapidlyto every continent except Australia." 8

The complicated history of horse evolution was sorted outby George Gaylord Simpson, the American paleontologist who, in1949, could claim, "The record has demonstrated that evolutionis not some overall cosmic influence that has been changing allliving things in a regular way throughout the periods of the earth'shistory."9

Spencer, of course, was a firm believer in "some overallcosmic influence" which propelled not just the evolution of life, butincluded the evolution of the stars and the arts as well. This beliefjust does not fit the record.

Stephen J.Gould has shown why people have been confusedabout the apparent movement towards complexity.1"His argumentis that, if you start with single cell creatures in a space of possibili-ties, then there is much more room in the direction of complexitythan in the direction of less complexity. Nonetheless, viruses whichare simpler than single-cell creatures are among the most successfulcreatures around today.

What Darwin actually wrote was that, if complexity exists,there is only one way that it could have arisen-through a series ofgradual changes with selection at each step-"If it could be demon-strated that any complex organ existed which could not possiblyhave been formed by numerous successive, slight modifications,my theory would absolutely break down." 11This is not the sameas saying that things must become more complex. It is not the sameas saying that all change must be gradual. And it doesn't even say

8 "AHandbooknEvolution,oAccompanynExhibition"TheBritishMuseumNaturalistory]959, ndenlargeddition).

9 G.G.Simpson, heMeaningfEvolution(NewHaven, T: aleUniversityress,1949).

10 StephenayGould, ifesGrandeur(London:onathanape, 996)PublishedinUSA yHarmony,ewYorks FullHouse).

11 Charles arwin, heOriginfSpecies yMeans fNaturalelectionLondon:.Murray,859 6th dition, 1371),8and402.

DesignIssues: Volume 0, Number Autumn 004


8/17

thatthingshave to change.Onthe contrary,Darwinwas aware hat,given stableconditionsof life, things could stay the same forlongperiods of time. In his own words, "Anumberof species... mightremain or a long periodunchanged,whilst within the sameperiod,severalof thesespecies by migratingnto new countriesandcominginto competitionwith foreignassociatesmightbecomemodified."'"

It follows that Spencerianevolution differs from Darwinianevolutionin two majorways:the formerhas botha forceforchangeand a direction, the latter has neither-or at least it has nothingcomparable o a law of gravity.Inbiology,Darwinianchangedoeshave trends,pressures,andso on,but mainlywithin a limited timespan (remembering hat a "limited" ime in biology may be thou-sands of years).Similarly, ome writers referto trajectories f tech-nological change,but these are not like trajectoriesn physics;theyareunpredictable vera longtime scale.Manyof thechanges nbothbiologyandtechnologyseem to be the result of accidents.Butif weneglect the possibilityof the existence of some patternsof change,we end up with the minimaliststancethateverythingthathappensis contingenton circumstanceshatare neverrepeated.

As statedby Douglas Adams (the authorof Hitchhikers uidetotheGalaxy):

Anything thathappens,happens,anythingthat in happeningcausessomethingelse tohappen,causessomethingelse to happen,andanythingthatin happeningcauses itself to happenagain, happens again.12

Naturalselection ies somewherebetweenthe extremesof a progres-sive forceand the absence of anything other than "ifit happens ithappens." There is, however another alternative-Lamarckianevolution-which has a direction progress),but replacesa "force"of natureby striving.It is obviously comforting o some people tobelieve that:(a) the world is getting better,and (b),thattheir owneffortsplay a small part in this process.Such people would be athome with a Lamarckianheory of evolutionarychange in design.The LamarckianAlternativeWehave now reachedthe point where we can returnto Ozcan'sinterestingquestion-what would CharlesDarwinhave made of carsand kettles-or ratherkettles andbicycles, since cars had notbeeninvented. His firststeps might have been to realize the importanceof ideas,and to decide that deas aboutartifacts anbe calleddesign.It is just possible thathe also might have thought that changes indesign were Lamarckian.Towards he end of his real life, Darwincamevery close to acceptingthe Lamarckiandea of inheritanceofacquired haracteristics.fhe hadthought hatchanges n kettlesandbicycleswereLamarckian,e would havebeensupportedby several

12 D.N.Adams, he almon fDoubt(London:acmillan,002), 9.

10 Design ssues:Volume0,Number Autumn004


9/17

modern writers. For example, Nobel Prize-winning biologist PeterMedawar, discussed the "evolution" of tools or instruments which:

undergo a slow systematic secular change of a kind whichit is perfectly possible to describe as an "evolution" ...provided of course one realises that it is the design of theseinstruments that undergoes the evolutionary change andnot the instruments themselves, except in a quite unneces-sarily figurative sense.l3a

Elsewhere, Medawar claimed that this kind of evolution isLamarckian and not Darwinian because-"It embodies a learningprocess." 13b Medawar was emotionally in favor of Lamarck becausehe wanted to believe that striving and learning achieved somepermanent improvement.Lamarckian ideas can be summarized as:

1 Striving to meet a need leads to greater use.2 Greater use leads to improvement.3 Improvements can be passed on-inheritance of acquired

characteristics.Lamarckian change is more than the inheritance of acquired char-acteristics, the muscles of the blacksmith being the classic case. Onits own, the ingredient of acquired characteristics does not work. AsHelena Cronin points out, why just the muscles, why not the badback and the burnt hands, and what about the blacksmith's daugh-ters? 14 The answer to why just the muscles being passed on is theingredient of striving, and people who are in favor of striving tendto wish that Lamarck was right.

Another reason for liking Lamarckian deas is that anything isbetter than leaving things to "blind chance." This reason was appeal-ing to many people including H. G. Cannon who, as a LamarckistProfessor of Zoology at Manchester University, made life difficultfor zoology students when I was a chemistry student there. In thepreface to his book, Cannon states, "If I can make it understood thatevolution represents a continuous succession of amazingly efficientthings that work, and not an incredible series of successful 'treblechances,' then I shall feel that I have been justified, for this I amcertain is the only way we shall escape from the arid conditions ofmodern genetical theory."15

The idea that Darwinian change is just "chance" is wrong,and the idea that biological change could be Lamarckian hasbeen convincingly demolished by Richard Dawkins, who states,"Lamarckismis not just something that might be; it actually couldn'tbe... the theory is in principle incapable of explaining the evolutionof serious adaptive complexity not just on this earth but anywherein the universe." 16

Dawkins points out that not all acquired characteristics are"improvements." The thing that separates changes that are improve-

13a.P.B.andJ. S.Medawar,heLifeScience,Chapter(London:ildwoodHouse, 977), 2.

13b P.B.andJ. S.Medawar,ristotleoZoos:APhilosophicalictionaryfBiology(Cambridge:arvardniversityPress, 983), 7. See alsoP.B.Medawar,"TechnologyndEvolution"The rankNelsonDoubledayectures,ewYork,1973).

14 Helena ronin,heAntand hePeacock:AltruismndSexual electionromDarwinoToday(Cambridge:ambridgeUniversityress, 991).

15 H.G.Cannon,heEvolutionfLivingThingsMancester:anchesterUniversityress 958),x.

16 Richardawkins,heBlindWatchmaker(London:ongmans,986), 88.



10/17

ments from those that are not is a selective environment. (We haveto remember that "environment" is not just the weather and stuffthat "Greens"worry about. For a particular gene, all the other genesare part of its environment. It has to FIT-to Function In Time andFit In Too.)

Discussing behavior, Dawkins said, "Suppose the skillsacquired during life by animals could be translated into DNA andget passed on. They would be one jump ahead, and evolution wouldbe speeded up." However, "This all presupposes that the changesin behaviour that we call learning are, indeed, improvements.Why should they necessarily be improvements?... there must be aDarwinian underpinning to ensure that acquired characteristics areadvantageous." 16

In other words, while evolution might happen somewhere inthe universe in a manner which involves striving and the inheritanceof acquired characteristics, such a system would not exhibit adapta-tion; it would not exhibit the appearance of design. The main reasonfor this is that there is no way of knowing what to strive for.

Striving has to be seen as a necessary but insufficient factorin Darwinian change. Any animal that inherited a lazy dispositionwould have reduced chances of passing on such a disposition (thehuman animal being an exception, of course). Animals have to spendtheir lives striving to keep up with the demands of the four Fs-feed-ing, fighting, fleeing, and the other F (in my view, the four Fs of thelimbic nervous system need the addition of a fifth-fun). This isjust a base line; they need a competitive edge if they are going tosurvive and replicate. The nature of the competitive edge is selected(not caused) by the animals' surroundings including other animals,sources of energy, and sources of danger.

So it is with human design. Ideas compete for resources, firstwithin the head of an individual designer, then within an organiza-tion, and then in the selective world of purchasers and users. Butsurely human design is different: humans can imagine somethingthat does not exist and organize resources to make it exist. This is thenature of striving, and Dawkins's objection to Lamarck in biologyalso applies to human design. The problem is that the best designerin the world has no way of knowing what the future will bring.Assumptions about what would make an improvement are notori-ous for coming up against unanticipated obstacles.

Changes in the environment can lead to the results of striv-ing becoming redundant. What happened to the large muscles ofthe blacksmith when no one wanted blacksmiths anymore? Once, afaster airplane could be assumed to constitute an improvement. Thencame the Concord, made possible only by massive expenditures bythe French and UK governments. Similarly, the designers and engi-neers who developed the Hovercraft thought they were strivingfor an improved form of transport, aimed initially at the need fora transport system that could cover both land and water. However,

12 Designssues:Volume0, Number Autumn004


11/17

the helicopterfilled thatniche and the secondaryaimof a "better"way of travelingover waterwas defeatedby otheradvances nwatertransport uch as thehydrofoil.

Sixdifferent irms who triedto make and sell theHovercraftallhad to give up the attempt.Eventhoughall thecosts of develop-ing a workingHovercraftwere paid forby the BritishGovernment,it was not possible to make a profit from the manufactureand saleof somethingthatpeople did notwant.'7

In otherwords,theLamarckianlternativewhichsounds likea description of human design-things gettingbetterthroughthestrivingof individuals-in fact does not work. Ifwe wish to discussdesignevolution,we have to considerDarwiniannaturalselection.NaturalSelection Outside BiologyAlthough he did not specificallymentionkettlesand bicycles,wedo not have to do muchguessingto have a good ideawhat Darwinthoughtaboutchange n non-biological ystems.In his secondgreatbook, TheDescentofMan and Selection n Relation o Sex,he speculatedabouttheapplicationof natural election n areasoutside of biology.Forexample, on language he stated, "Thesurvivalorpreservationof certainfavouredwords in the struggle for existence is naturalselection."Darwinquoteda writer n Naturen 1870who wrote,"Astruggle or life is constantlygoingon amongst hewordsandgram-matical forms in each language. Thebetter,the shorter, he easierforms areconstantlygaining the upper hand, and they owe theirsuccess to theirown inherentvirtue."As Darwinpointedout,thereis moreto "success" han"inherent irtue";anguagedoes notneces-sarilyprogressn thedirectionofbeingmorevirtuous.He suggested,"Merenoveltyandfashionmaybe added forthere s in themindofmana strong ove forslightchanges n all things." 8

Darwin was well aware of the importance of mind. Hesuggested that a sophisticated language requires a sophisticatedmind, andthe onlyway thatcould havehappened was by whathecalled "correlation f parts,"a term to describe how two differentthingschangedslowly togetherso thatthey could keep in step. Hehad similarviews on technicalchangeor invention:

If some one manin a tribe,moresagaciousthanthe others,invented a new snareorweapon,orothermeansof attackordefense,theplainestself interestwithout the assis-tanceof muchreasoningpower,would prompttheothermembers o imitatehim and allwould thusprofit.Thehabitualpracticeof eachnew artmust likewise in someslightdegreestrengthen he intellect.18

At firstsight,the natureof the selectionsystemwithinwhich ideascompete might be seen as being very differentfromthe selectionsysteminbiology,butDarwinhadtwo importantheorieswhicharerelevanthere.Thefirstof these is unconsciousselection.Becausewe

17 P.S.Johnson.The evelopmentofHovercraft,"hree anks eview(December974).

18 Charles arwin, heDescent fManandSelectionnRelationoSex(London:JohnMurray,871 2ndEdition,883]),129and91.



12/17

have somecontrolovertheselectionsystem, t mightbe thought hattheevolutionof kettlesandbicycleswas a formof artificial election,like animalsbeing bredby humans.However,Darwinwas awarethat even artificial selection did not proceed in a totally rationalfashion. He gives the example of two flocks which started out asdivisions of the same flockof Leicestersheep but, over fifty years,diverged from each other in an unpredictablemanner to such anextent thattheyhad "theappearance f being quitedifferentvariet-ies." He called thisphenomenon"unconscious election."19

Governmentslike to think that they make quite consciousdecisions to supportsome things and discourageothers,which ofcoursethey do but such decisionsmayhave "unconscious" ffects.The"rules"of the competitionbetweendesign ideas can be altereddeliberatelyby taxing some things and subsidizing others.Somethings can be made illegal, while awards may be given for otherkinds of things.Theproblemis that there is still the need for whatDawkins calls "aDarwinianunderpinning"becausegovernmentsdo not reallyknow what to support and what to discourage,andbecause of the unexpectedeffects of "unconsciousselection."TheBritishgovernmenthassupportedHovercraft ndhydrogenbombs.It supported arger amiliesby offeringchild benefits.It has bannedcannabisandworkingfor the governmentafteragesixty-five(apartfrom judges and prime ministers). When the contraceptive pillwas introduced in the UK,medical treatmentwas free under theNationalHealthService,butprescriptionsor thepillhad to be paidfor.Today,here s a charge or mostprescriptions, ut thepill is oneof the exceptions; t is free.

As asociety,we can use reasonto attempt o makeimprove-ments,but therealwaysis uncertainty boutoutcomesso we still areleftwith a Darwiniannaturalselectionsystemunderpinningouref-forts.A good exampleof this can be foundin England,where a gov-ernment-fundedcull of badgerswas carriedout to reduce the inci-dence of tuberculosis n cows that cancatchTBfrom infectedbad-gers. Badgersare socialanimals hat live in smallgroupsanddo nottravelvery much as long as theyhave food andcompany.Attempt-ing to kill the badgers destroyedthe groups and left lone badgersroaming hecountryside.Thisapparently ed to a twenty-sevenper-cent increase in bovine TB in areaswhere badgers had been shot,compared o controlareaswhereno shootinghad beenallowed.20

Even when we arevery surethatsome changewould be forthebetter,suchchangewill still have the unexpectedside effectsofDarwin'sunconsciousselection.

Darwin called the second of his theories that concern theselectionsystem"sexualselection":

Thenests of hummingbirdsandthe playingpassages ofbowerbirds aretastefullyornamentedwith gaily coloredobjects;andthis shows thatthey mustreceivesome kind ofpleasurefrom the sight of such things.2,

19 Charles arwin, heOriginfSpecies yMeans fNaturalelection,(59), 5.20 "Badgerilling ed oRise nTB," he

Guardian,November,2003): .21 Charles arwin, heDescent fMan nd

SelectionnRelationoSex(1883),2.14 DesignIssues: Volume 0, Number Autumn2004


13/17

Darwinsaw the main roleof pleasingsightsandsounds tobe sexualattraction, nd he was fascinatedby thepeacock's ail.Clearly,a tailwhich is large and heavy has no advantagein survival terms-itrequiresenergyandit advertises ts presenceto anypassingpreda-tor.Darwin'sexplanationwas thatbasically hefemalepeacockshada preference orelaborateails,and the evidencefor thiswas thatthepeacocktailsareat theirpeakduringthematingseason.He called this phenomenon, "sexual selection." Anotherexamplewas theArguspheasant.Referring o theArguspheasant,Darwinstated:

He who thinksthatthe malewas createdas he now existsmust admit thatthegreatplumes, whichprevent thewingsfrombeingused forflight and which aredisplayedduringcourtshipand atno other timein a mannerquitepeculiarto this one species,were given to him as an ornament. fso, he must likewiseadmit that the female was createdandendowed with the capacityof appreciating uchornaments.I differonly in the conviction hat the maleArguspheasantacquiredhis beautygradually, hroughthepreferenceof thefemalesduringmanygenerations.22

So I amsurethat Darwinwould havebeenhappyto see kettles andbicycles in terms of the evolution of ideas, and thathe would nothaveseen such a processasbeingparticularly rogressive.Hewouldhavefoundroom forfashion,"astrong ove forslightchanges"and"1unconscious"esign.Hewould havedrawnananalogywithsexualselection, and he also would have been happy with the notion ofimitationas one way in whichideas arespreadunderthe influenceof "theplainest self interestwithoutthe assistanceof muchreason-ing power."

So farI have tried to establish(1)thatideasof evolutionthatexist in the design literatureareconfused or pre-Darwinianand,should be consignedto thewastebasket; 2)theyshould be replacedby a nonprogressiveDarwinism; nd(3) that theformof Darwinismthat is needed to makesense of change n designis the evolutionofideas.

Inthe1930s,biologyachieveda synthesisof the ideasof Dar-win with the ideas of genetics andthemutationof genes to produceneo-Darwinism.Genesprovidedananswerto theproblemof repli-cation.Mutationprovidedananswerto theproblemof thesourceofnew varietieswithoutwhichnaturalselectioncomesto a halt.A neo-Darwinianview of design changeis naturalselectionplus memes, theircompetition,their modes of transfer,and theirtransformation;.e.,memetics.2 Ibid., 16.



14/17

Towards he Memetics of DesignDawkins's memes which, in this context, are design ideas that can bereplicated do not have to wait very long for replication to take place.They speed up the old genetic form of Darwinian change, but theevolution of design ideas is still Darwinian because ideas about whatto strive for are in competition for scarce resources to turn them intomanufactured realities. There are no basic principles telling us howone group of designed objects is superseded by another.The processessentially is unpredictable. There is no law of selection "to propelthings in the direction of progress."Selection is blind because there isno way of knowing what happens next. Nonetheless, we keep trying.If we stop striving for improvement, we have stopped being human,but we should not be surprised if our efforts sometimes fail. Oncethis apparently gloomy view is absorbed, it can be put to work.

What might be called "Darwinian design under the influenceof natural selection" was first used to make money by the Germandyestuffs industry in the nineteenth century. Teams of skilledsynthetic organic chemists were employed to make novel, coloredchemical compounds. Since there was no way of knowing which ofthese would make useful dyes, the new compounds were tested ina dye house where most were found to be useless, but some wereselected for further chemical modification in the hope of improvingthem. By 1910, it was calculated that ten thousand new compoundshad to be tested to find one new commercial dye, but the profits fromthe one, successful dye were much greaterthan the cost of producingthe ten thousand.

I find it reasonable to call such a process Darwinian, butDarwinian processes can have very surprising side effects. Thedemands of the German chemical industry for university trainedorganic synthesizers to make the new compounds led to the inven-tion of a "juniordoctorate"-the Ph.D.-which would take less timeto achieve than the traditional higher doctorate, the D.Sc. The Ph.D.spread from Germany to the U.S., and then around the world sothat today's potential academics in any subject have to obtain anacademic qualification which was invented for the German chemicalindustry. I do not detect any sign of progress in this particular step inthe evolution of education. The "meme" of a Ph.D. has been remark-ably successful in propagating itself and, like the peacock's tail, it hasprospered because it is "fancied" and not because it is "better."

However, there is a danger in replacing Spencer or Lamarckby memetics, and that is the replacing of one confused way of think-ing with another. The achievements of memetics so far have not beenimpressive. Elsewhere, I have argued that, if memetics is to develop,it needs to do three things. The first is to move away from its concen-tration on imitation and epidemiology. The second is to realize thatthinking of memes as "units" is not helpful. They are "patterns."

16 Design ssues:Volume0, Number Autumn004


15/17

The third is to recognize that there are different types of memeswith different methods of transmission. These are "recipemes,""selectemes," nd "explanemes."3

Recipemesare transmittabledeas about how to do things-recipe ideas.

Selectemesare deasaboutwhat sort of thing you want to do.Theyare involved in makingdecisions between alternatives.Theyprovidemotivation; hey are values.

Any designer working for a client has a set of ideas aboutwhat the client wants.Theyalso have ideas aboutthe marketplace,aboutfashion,andabout hesortsof designsthattheirpeers approveof.These deas arenot workedout like a physicsequation.Theyforma "pattern"n the mind, what MariaAbu-Rishacalls a "patternofneed."I think thatthis is a patternof selectemes.

At the same time, designers have other groups of ideas ofthingsthat arepossible, deasabouthow to makethings-recipemes.These forma patternof possibilitieswhich arecomparedwith thepatternof need untilthereis a "click"-a fitbetweentheselectemesand the recipemes. MariaAbu-Rishacalls this click "purposivepattern recognition" (PPR).24 t is purposive because the designerknowswhatto do next.

The sameconceptsof selectemesand recipemescanbe usedwhen thinkingabout how design changesover long time periods.Theyarenot restricted o an accountof the here and now of a specificact of designing.Inthe sameway, genes areused to describewhathappens at the conceptionof an individual life, and they also areused to discuss how things changeover millionsofyears.Bothgenesandmemes areevolutionaryreplicators.

The thirdtype of meme, the "explaneme,"must be addedbecauseof the humanpropensity o ask"why?"As long as humanshave had a language, they have told stories, and good stories getreplicated. f someone discoversa new recipe,peoplewill askwhyit worksas well as how it works.Explanemesare deas thatprovidethe basis for answering "why" questions. They range in sophisti-cation fromsimple storiesto complexmathematicalconcepts, buttheyhave two thingsin common, heyofferanexplanationandtheyneed a languageto be transmitted.They differ n this fromthe othermemes which sometimes canbe transmittedby imitation withoutformal anguage.

Explanemes forman essential partof the discussion aboutDarwiniandesignchangebecause of theclaimthat humanrational-ity, science, and mathematicalengineeringmakes modern designchange different from the days of craft design when people didnot know what they were doing (they still had stories though). Anessentialpartof the claim thatdesign change s Darwinian and notLamarckian) s the Dawkins "knockout"that all that rationalitycounts forlittle if we do not know what is going to be "better." nfact,our ideas of improvementare themselvessubject o Darwinian

23 J. Langrish, Different ypesofMemes:Recipemes,Selectemes, andExplanemes,"oumal of Memetics3(1999). www.cpm.mmu.ac.uk/jom-emit/1999/vol3/langrishz.html)

24 MariaAbu-Risha, Purposive atternRecognition,"Ph.D. hesis, De MontfortUniversity,999).

Design ssues:Volume0,Number Autumn004 17


16/17

change.Selectemesarein competitionwith other selectemesand,atdifferent imes and in differentplaces,some aremoresuccessfulatbeingreplicated hanothers.Returning to FortyAs stated above, Fortyrejectsboth "greatmen" and evolution asexplanationsfor changes in design. So what does he put in theirplace?The short answer is "ideas"-but what sort of ideas? Nearthe startof his book,he claims:

Everyproductto be successful,must incorporatehe ideasthatwill make it marketable, nd theparticularaskofdesign is to bringaboutthe conjunctionbetweensuch ideasand the availablemeans of production.25

Forty,of course,is aware that"the availablemeans of production"are themselvesdesignedandsubject o change,but he does not likethe ideathattechnologicalchange"causes"designchange.He callsthis "themechanical allacy."26

As an example, Forty shows how the mechanization ofsewing in theUnited States and in VictorianEnglandwas followedby a fashion for heavily trimmed dresses (i.e., lots of elaborateextra materialsewn on to the basicdress).Thisadditionalsewingwas achieved at little orno extracost to the purchaserof the dressbecause hesewingmachinecouldsew muchmorecheaply hanhadpreviously beenthe case.

Fortyresists heconclusion hatthesewingmachine"caused"the fashionchangeby pointingout thatsewingmachinescouldhavebeen used to reduce he hoursworkedby the machineoperatoror topay the workersmore. He concludes:

Thus the ultimatecause of the fashionfor heavily trimmeddresses was not now the sewing machine tself, but its usewithina capitalist ystem of manufacture.26

Now, the conceptof an "ultimatecause" s another dea from oldertheories of change.Darwinianchangehas few, if any,events thatmightbe labeled "ultimate."

In a Darwiniansystem, the recipemes(e.g.,sewing machinetechnology)have to FIT nto an environment of selectemeswhichincludes ideas of desirabilityheld by thosewho put up the moneyforthetechnology,as well as ideasof desirabilityheldby those whobuy theproductsof the technology.

Fortyand I,however,do agreeon one thing,and thatis theimportanceof changingideas. Tome, the history of design is thehistory of ideas-ideas about how to make things which I like tocall recipemes,and ideas about what sortof things to make whichI like to call selectemes. People make choicesbetween competingideas, and they sometimesuse another kind of idea to justify theirchoice.Suchexplanatory ustifications remy explanemes.All these

25 Adrianorty, bjectsfDesire: esignandSociety ince1750. .

26 Ibid., 1.18 Design ssues:Volume0, Number Autumn004


17/17

ideas and their nteractions an be said to evolve. These deas are thememes of design, and I would hope that a modernDarwinwouldagree that their evolutionis aDarwinianprocess, nvolving compe-tition for resources to ensuretheir survivalbut lacking long-termpredictability or two reasons:first, the "rules"of the competitionkeep changingand, secondly, uccess n being replicated s subject ochanceandwhimsy.Spencerianprogress s nowhereto be seen,andshould be consignedto the waste basket of rejected xplanemes.

[paper] darwinian design thememetic evolution

Documents