Multiple   Independent   Inventions   of   a   Non ‐ functional   Technology   Combinatorial   Descriptive   Names   in   Botany ,   1640 ‐ 1830  

Historians and sociologists of science usually discuss multiple independent inventions or multiple independent discoveries in terms of priority disputes among the inventors. But what should we make of the multiple invention of a technology that not only gave rise to very few priority disputes, but never worked and was rejected by each inventor’s contemporaries as soon as it was made public? This paper examines seven such situations in the history of botany. I devote particular attention to the inventors’ cultural and educational backgrounds, in particular, the scholastic education most of them shared, through which they would have become familiar with Llullian combinatorics and the mnemonic names used to distinguish syllogistic moods. I also examine their conceptions of the roles of nomenclature in botany, their assumptions about how memory works, their awareness of other similar efforts, and their contemporaries’ reactions to their proposals. Finally, I reflect on the impacts that a consideration of multiple independent inventions of failed technologies may have on current approaches to the history and sociology of science.

previous research into multiples, writing, "The innovations became virtually inevitableascertainkindsofknowledgeaccumulatedintheculturalheritageand as social developments directed the attention of investigators to particular problems"(Merton1961, 475).Mertonwasconcernedwithdemonstratingthe reality,ubiquityandimportanceofmultiplesinthesociologyofscience.Hiswork andtheworkofothersafterhimwhowroteonmultiplesandtheircauses(such asKuhn1959;Simonton1986and2003,LambandEaston1984)thereforefocus on the roles of social interactions among scientists in scientific creativity, scientific "genius," and priority disputes concerning successful research programs.They were concerned about what multiple discoveries of natural phenomenamightimplyintermsoftechnologicaldeterminism.Theywerenot interestedinmultipleindependentfailures.
One of the few sociologists of science who did mention failures, Augustine Brannigan,evenwentsofarastosaythat"wewouldfinditextremelycuriousor bizarre to find historians and sociologists of science applying the cognitive explanation of discovery to scientific failures" because "a theory of discovery shouldnot focus onhow ideascame intothemind, orhowtheyevolveasthe culture matures, but how they are defined as discoveries."For Brannigan, restrictinganalysistothesocialnegotiationofwhatcountsasadiscoverywasa way to avoid the slippery slope of technological determinism by staying away from teleological interpretations of how and why given discoveries occurred whentheydid (Brannigan1981,40,152).
Unlike Brannigan, I believe that multiple independent inventions dismissed as failures by the inventors' peers reveal important things about the social aspects of science.In fact, there is nothing quite like a multiple, independent failure-reinvention of a dud-to highlight the assumptions inherent in a researcher's cultural heritage.Multiple duds are background assumptions brought to the foreground and writ large.They illustrate attempts to solve common problems with inappropriate tools, or to combine common ideas in unsuccessful ways.Their existence testifies to an unequal distribution of knowledge among scientific practitioners, since inventors of multiple duds 2 are not only unaware of the invention of previous, similar approaches, but also of their failures: even those inspired by previous duds, or who plagiarize earlier duds,misjudgetheirproposals'potentialforsuccess.Theapproachtakenbythe dud-makers was just one of many different tacks taken to improve specimen identificationandbotanicalnomenclaturefromthemid-17 th centurytothemid-19 th century.I have discussed many of the other methods that contemporary botanists explored in other publications (Scharf 2007;Müller-Wille and Scharf submitted 2008;Scharf 2008).My research into pre-Darwinian botanical classificationandnomenclatureschemeshasalsoledmetobelieve,alsounlike 2 I refer to failed inventions as "duds" in this paper because the word "dud" is shorter and catchierthan"inventionjudgedtohavebeenineffectualbytheinventor'speers,"ratherthanto trivialize the efforts that each inventor put into devising a solution to the technical problems facedbybotanistsofhistime.
Brannigan, that addressing the teleological aspects of successful scientific research programs head-on may be a fruitful approach to understanding scientific discovery.The existence of multiple duds, for example, supports an evolutionary epistemology for scientific development: duds are attractive enoughtobeinventedmultipletimes,buttheyarealsorepeatedlyrejectedas flawed.Dudscan,infact,beseenas"deadends"intheevolutionofideas.Inthis case, multiple duds are inevitable by-products of the "constrained stochastic behavior"ofscientificcreativity,inparticular,theindependentactionsofpeople trying to solve similar problems in similar ways, with similar, though inappropriate, tools (Simonton 2003 and2004, particularly pages 91 and184).An evolutionary epistemology of invention can account for the erratic and yet sometimesseemingly teleological development of successful scientific research programs, as well as the occurrence of multiple duds.It may well be that an evolutionary epistemology of invention is the long-sought middle ground between the chaos of many unconnected "micro" case studies and the overly deterministic slant that longue durée or "macro" studies tend to take (Misa 1994,119).
InthispaperIdescribethemultiplereinventionofatechniqueforreferring toplantsby"names"formedfromcombinationsofletters,eachofwhichcodes for a particular plant feature.I examine in turn the common problems each of theinventorstriedtosolve,thecommonapproachesorsolutionstheydevised, the shared assumptions they held, and why these assumptions caused the proposed solutions to fail.Finally, I explore the possible implications these multiplereinventionshavefortypicalapproachestothehistoryofscience.

ACommonProblem:LargeNumbersofPlantSpeciesandEvenLargerNumbersof NamesforThem
One problem that naturalists studying plants increasingly faced during the 17 th , 18 th , and 19 th centuries was how to keep track of large numbers of plant species.TheBauhinbrothershaddescribedmorethan6,000kindsinthe1620s.This number ballooned to over 18,000 in the works of John Ray only 80 years later, and the rate of discovery continued to increase.Botanists recognized in the mid-17 th century that there were far too many known kinds of plants to allow for the memorization of their names, let alone information about each one.Andyet,therehadtobesomewaytodifferentiateeachplantfromothers.To make matters worse, multiple names-synonyms-often referred to the sameplant.Itwaswidelyrecognizedthateliminatingsynonymswoulddomany great things for botany.It would cut down the number of plant names in circulation, clear up priority disputes, consolidate partial descriptions so that poorly known species could be understood better, and reduce the amount of time spent tracking names-something that everyone agreed was a tedious bore.
Synonyms come about for practical reasons.First, not everyone becomes awareoftheexistenceofagivenplantatonce,and,second,identificationsare notalwaysprecise. 3Onewaytosolvethe firstproblemwouldbetosomehow broadcast the name and a description of a plant around the world once it had been described and named.But during the 18 th century, there was no central registryfornames,andnoformallycodifiedrulesaboutwhocouldnamewhat, or how complete partial specimens had to be, etc.Furthermore, the transit of books,lettersandjournalstothetotalityofthecommunityofnaturalists,then inhabiting every continent save Antarctica, was slow and erratic.Eventually, naturalists'frustrationwiththeseveryissuesledthemtoestablishcodifiedrules governing nomenclature in the 19 th century, later formalized in the late 19 th centuryastheICBN(forplants)andtheICZN(foranimals).Modifiedversionsof these codes are in effect today.While the codes are far from perfect and will neversatisfyeveryone,theyhavebroughtagreatdealofordertothenamingof organisms.This makes the study of the state of natural history prior to the establishment of the codes particularly enjoyable.Until the Strickland rules of 1842, natural history was an especially vibrant free-for-all of interesting ideas and experiments to bring synonymy and related problems under control (McOuat1996).
Clear communication about plants was a common need; there were also commonlimitationsandtools.

CommonTools
Botanists in the 18 th and early 19 th centuries (and even earlier) found it convenient to describe plants in terms of obvious, preferably external morphological features.In practice, these were most often the numbers and positionsofflowerparts,aswellasthegeneralexternalshapeofplants.These approaches had many practical benefits, since the numbers and positions of flowerparts,and"lifeform"ofplantspecies-tree,bush,vine,herb,andsoonare generally their most constant morphological attributes.Other plant parts vary much more from individual to individual within the same species-even within cuttings taken from the same source.This is one reason why it is much more difficult to distinguish different kinds of plants than different kinds of animals: most animals have fixed numbers of parts and are subject to less environment-inducedvariation(phenotypicplasticity).Botanistswhowishedto telldifferentplantkindsaparthadtodevoteagreatdealofefforttofindingout whicharrangementsofpartsarethemostconstantwithinplantgroups.George LouisLeclerc,ComtedeBuffon(1707-1788),attributedthecontrastbetweenthe elaborateand reasonedclassificationsinbotany andthelagindevelopmentof similar schemes in zoology to this difference between plants and animals.EchoingBuffon,theFrenchbotanistLouis-Marie-AubertduPetit-Thouars(1758-3 These problems of definitions varying in precision and of patchy communication among scientistsareexactlythesameonesthatleadtomultiplediscoveriesinscienceingeneral. 1831) wrote in 1788 that systems and methods of classification were first and most extensively developed in botany rather than zoology because there are more plants than animals, plants resemble each other more than do animals, and plants have fewer obvious body parts upon which to base classifications thananimalshave (Buffon1749,tomeIIp.10;duPetit-Thouars1811,27-30). 4  Usingdescriptionsofplantsintermsoftheireasilyvisibleouterpartswasa matter of exigency.Other approaches to understanding plants did not develop fullyduringthe18 th orearly19 th century.Zoologistsinthelate18 th century,such asthecomparativeanatomistGeorgesCuvier(1769-1832),begantouseinternal anatomy to characterize animal groups with great success, as well as making headway toward understanding the functions of animal organs.Comparative anatomyandphysiology,however,turnedouttobefarlessusefulforstudying plants.Whilemanyanimalbodyfunctionsarelocalizedtoparticularorgansthat occur in particular places and in particular numbers, plant functions are more distributed throughout their bodies.Until microscopy became relatively cheap andwidespread,andcellfunctionswerebetterunderstood,plantanatomyand physiology were sufficiently alien to preclude causal explanations of plant growth,developmentandstructure(e.g.Duméril1806,x,andCandolle1813,59;Stevens1994,252note287,providesmorereferences).
This situation made the classification of plants according to the numbers, shapesandpositionsofpartsparticularlyusefulasaframeworkofinvestigation.It is not necessary to know what leaves or stamens do in order to notice their shapesandmodesofattachment,ortocountthem.Itwasahappycoincidence for 18 th century botanists that they understood the function of stamens and pistils,whichwerealsothemostdistinctiveandconstantflowerpartsintermsof shape,positionandnumber.
CarlLinnaeus(1708-1788),themostinfluentialbotanistofthe18 th century, usedthissituationtojustifytheappropriatenessofhissexualsystem-basedon stamenandpistilnumberandposition-forclassifyingplants (Müller-Wille1995, 45and2005,90-91).Thesexualsystemconvenientlyworkedforidentifyingboth live plants in the field and dried herbarium specimens, so long as the flowers were visible.However, flowers are not visible for all stages of a plant's life.Linnaeus, like a number of other 18 th -century botanists, also strove to classify plantstakingmoreoftheirfeaturesintoconsiderationthanjusttheflowers.His 4 Thenumberofdescribedplantspeciesexceededthenumberofdescribedinsectspeciesuntil thelastdecadeofthe19 th century.Thepopularbeliefthatthereweremoreplantsthaninsects persisted at least into the 1830s (e. g.Henderson 1832, 160), though the number of insect species estimatedtoexistbycompetententomologistsbegantoexceedtheestimatednumber ofplantspeciesinthesecondquarterofthe18 th century(Scharf2007,180).Wenowknowthat flowering plants are an evolutionarily young group composed of members much more closely related to each other and more similar in appearance to each other, on the whole, than, for instance,animalsindifferentphyla.
"fragments of a natural method," first published in Classes plantarum (1738), keptspecieshesawasrelatedtoeachothertogetherinnaturalgenera.Because thesegenerawereacceptedasnaturalandrelativelystable,Linnaeusreferredto themasthe"commoncurrency"ofbotany.Fulldescriptionsofnaturalgroupsof plants, what Linnaeus called their "natural character," could work in any botanical system, regardless of whether the overall classification or nomenclatural schemes were completely different from each other (Linnaeus 2003,142 §189).Linnaeusclearlyunderstoodthevalueofkeepingclassification andnomenclatureseparate(e.g.Müller-Wille2007,542,discussing,inparticular, a passage from the introduction to the first edition of Genera plantarum [Linnaeus 1737[Linnaeus , lectori 2003, 141-143], 141-143]).Keeping description separate from nomenclature allowed Linnaeus to use short and convenient binomial nomenclature and to avoid and advise against "generic names 1 ½ feet long, those that are difficult to pronounce" and, in particular, "words that contain more than 12 letters" (Linnaeus 2003, 214).Linnaeus's suggested naming conventions were practical.Botanists all over Europe began adopting them in the1750s,andbythe1780s,theywerecommonplace.Linnaeus'sseparationof nomenclature from the descriptions used in plant classification is one reason whythebinomialnomenclaturehehelpedpopularizeisstillstandard,centuries afterhissexualsystemofclassificationbecameobsolete.
Separating description from prescription and from naming, however, has always been tricky.Many 17 th -, 18 th -, and 19 th -century botanists found the conceptofdescriptivenamesparticularlyattractive.Atleastsevendifferentmen living in six different countries and writing in four different languages independentlyinvented-orclaimedtohaveindependentlyinvented-variations on a technique for generating plant names using descriptions of the plants themselves.These descriptive naming schemes were meant to prevent synonyms from being formed in the first place.With this kind of system of nomenclature,aplantwouldgetonlyonename,regardlessofwhowasnaming it.Descriptive names would also make botanical nomenclature less arbitrary thanithadeverbeen.
As to be expected, the particulars of these descriptive naming schemes varied from inventor to inventor and bore the imprints of trends popular at different times in the history of botany.These differences, along with biographical details, lend credence to the likelihood that their inventions were independent.Still, there were sufficient similarities among the inventions to meritgroupingthemtogetherasexpressionsofthesameidea.

CyprianKinner'sScheme(1645)
Kinner was a Silesian lawyer and physician.He lived most of his life during theThirtyYears'War.Fewdetailsofhislifeareavailable,thoughheisknownto havecomefromawealthynoblefamilyfromtheSilesiantownofBrieg,where he studied at the same school as Samuel Hartlib.He travelled extensively, and hadmanyintellectualcorrespondents.Hewaslonginterestedineducationand language reform, and had worked for a time in Transylvania with Johann Heinrich Alsted (1588-1638), an encyclopaedist.After he lost his family and fortunewhentheImperialarmyinvadedSilesia,"heofferedhisservicein1644 to the Czech reformer-in-exile Jan Amos Comenius (Komenský)" (1592-1670), BishopofBrethrenandastudentofAlsted's,whohiredhim1645ascollaborator on the revision of a language textbook. 5It was around this time that he developedhisschemefornamingplants(DeMott1958).
Kinner'sletteroutlininghisideaisparaphrasedina1958articlebyBenjamin DeMott.Afewyearsbeforehewrotetheletter,Kinnerhadexplained, inaperiodwhenhewasconsideringwaysofhelpingstudentsof botany,he,too,hadthoughtofdevisingtechnicalwords(voculas technicas).Hisnotionwastofashionthemsothateveryletteror syllableinthemwouldhaveaspecificmeaning.Consonantsinthe first syllable would denote primary and secondary qualities … vowelswoulddenotethedegreeofqualities.Thesecondsyllable would express the peculiar power of the plant-curative, preservative,nutritive,orthelike.Thethirdsyllablewouldsignify evenmoreparticulardetails,asforexamplewhenandwherethe plant grows and how it is gathered.Some letters would necessarily be repeated in different syllables (the Latin alphabet notprovidingenoughlettersforalltheinformationthatwouldbe expressed in the word), but the repeated letters could be arrangedsothattheywouldhavedifferentsignificationaccording tothesyllableinwhichtheyappeared.Andthesyllablesmightbe made to vary in length from one to three letters, so that the meaningofletterswouldthendependupontheirplacewithinthe syllable: by such techniques the problem of repetition could be solved.Wilkins was a powerful man.He was friendly, somewhat dogmatic, bright, and he had very influential friends-for instance, he was married to Oliver Cromwell's sister (Maat 2004, 136).He was a founding member and the first secretary of the Royal Society, and he became the Bishop of Chester by 1668 (McMahon 2001, 240).He had been working on his own universal language basedonHebrewinthe1640s,butheneglectedthatsortofresearchuntilafter he met Scottish language reformer George Dalgarno (1616?-1687) and collaborated on his project (Maat 2004, 137-138).Although both men had similarideasofhowtocomeupwiththeindividualwordsforthings,theymade different assumptions about how memory works and how much it can retain.Theseassumptionsaffectedthedegreetowhichtheyexpectedthevocabularies of their languages to reflect their classifications and how they thought their languages would best be employed.By the end of 1657, Dalgarno and Wilkins hadafallingoutabouttheseissuesandwenttheirseparateways(Maat2004, 50-55;Lewis2007,88-99).
WilkinshadoriginallycompletedhisEssayin1666.InSeptemberofthatyear, the printer had almost finished with it when most of the printed sheets and a goodpartofthemanuscriptweredestroyedintheGreatFireofLondon (Maat 2004,135).
After the fire, Wilkins resolved to put the book together again, bigger and better than before.He enlisted his friends to help with different sections.The naturalistJohnRay(1627-1705)workedonthetablesofplants.
Wilkins had also provided for each of these location indicators to be expressed in words.Each genus was represented by a two-letter word.Each difference was indicated by adding a different consonant, and species were distinguished by the use of different vowels at the end.Using this notation, "18.1.4,"equivalentto"elephant,"couldalsobespelled"zibi"(Maat2004,158).Thenamesforplantsweretobeformedinananalogousway.
ChristopherPolhem'sSchemes (1739,1741) Christopher Polhem (d. 1751) was born in December 1661 on the island of Götland,justeastoftheSwedishmainlandintheBalticSea.6 Hewasborntoa merchant'sfamily,but,orphanedatayoungageandpenniless,hehadtowork as a farm hand to support himself.Friendly priests who spotted his talents helpededucatehimandgothimadmittedtotheUniversityofUppsalain1687at theageof25.Formuchofhislife,hewastheleadingmathematicianinSweden, andhewasalsooneofthefoundersoftheSwedishRoyalAcademyofSciences inthelate1730s.ApaintingofPolhem,madebyJohanHenrikScheffelin1741, nowseenbySwedesonthebackoftheir500-kronor note,showshimwearing theOrderoftheNorthStar.HelatersharedthishonourwithLinnaeus.Some time after 1739, his seventy-eighth year, Polhem approached the Swedish Royal Academy of Science with a manuscript on "Suggestions for Botanical Names."He wrote that he had been inspired by reading one of Linnaeus'sworks-possiblyClassesplantarum(1738)-todevotetimetohisown oldideaforauniversallanguagetodescribeplants.Thesystemheproposedhad different consonants at the beginning of words to indicate different kinds of substances."K"inparticularwastostandfortreesand"G"forgrassesandotherherbaceousplants.Theinitialconsonantdesignatingthesizeandrigidityofthe plantwastobefollowedbyavowel.Eachvowelwastosignifyadifferentsense thattheplantaffectedinthehumanobserver:"A"wouldstandforsight,"E"for hearing,"I"forsmell,"O"fortasteand"U"fortouch.Thelettersofthesenames couldbewrittenindifferenttypefaces,"[s]othatifgashouldexist,itnecessarily signifies an herb which sight alone can indicate or please, such as a beautiful flower, according to the approximate size that the size of the consonant will express" (Polhem,1954,346-347).Alongtheselines, ge can indicate an herb which man has heard about but not yet gained knowledge of its actual properties; gi indicates an herb that smells pleasant or appalling; go an herb that according to tasteisgoodenoughtoeat;andgu[aherb]which hasasoft or sharp feel etc.However, if a man wants to describe any herb's figureasfollows,thenthefirstsyllablemustbegå,becauseåisa compositeofaandu[such]astheFrenchwrite,andaccordingto a and u, vision and feel give the most reliable knowledge about everything...It happens sometimes that that one needs two vowels together, such as when an herb both smells and tastes good, which or consequentlyiswrittengioandtheirsizeasthesituationrequires (Polhem1954,348).
Tothethirdpositionwastobeassignedoneofthe"semi-vowels," such as s for the intellect, l for all operations by the tongue in general living, n for hands, m for feet orModus localis and r for thewholebody'slabourofinternalandexternalnature.Tofollow upthiswithexamplesbecomestooextensive,thereforeonlyone example here and there is included which can demonstrate the methodforeverythingelse;e.g.:gasdenotesaflowerwhichthe intellecthasmuchtoreasonabout,galrepresentsaflowerwhich hasmuchtobetalkedabout;gesrepresentsaflowerwhichmuch isheardaboutandmuchtalkedabout;ganaflowerwhichcanbe used for manufacturing and yarn colouring and which has been transported from another place; and gar a flower which serves our body as medicine or poison.The entry gis marks a sensible chemical mixture which causes good odours, etc., gil an herb whichacharlatancantalkaboutforhisownprofit,ginanherbof goodornastysmellwhich,however,servesforsomepreparations or manufacture, gim an herb that smells and which can serve to promote speed, such as hemp for ropes for riding, horse tackle, and rigging and sails, etc., gir an aromatic herb for medicinal purposes.The rest is easy to work out from this (Polhem 1954, 348-349).
But Polhem did not give up.Two years later, in 1741, he felt the need to revise his older manuscript after he saw that the Handlingar published a description of a collection of 100 plants that Linnaeus had discovered in Gothland, Öland and Småland.This new version of Polhem's was called "Suggestions for such Names for Herbs and Grasses thatwill in a Concise Way Point out their Virtues and Qualities in General."The main ideas of how alternating consonants and vowels assigned place value were to represent qualitiesthatobserversnoteinplantsremainedthesame.Healsoaddedsome commentsonhowtheuseofthree differentfontsandfivedifferentfontsizes could increase the number of plants it would be possible to describe with the system.Forinstance, bupointsoutanexternallyrecognizableherb,andcanberanked forbetterorworsewiththelargestvowelforcotton,followedby linen,andsohemp,thistleandnettle.
When these are divided into 3 kinds and each kind into 5 sizes, [there] appear 90 variations, and when the aforementioned 15,625 are multiplied with this, [there] occur 1,406,250.More differences should not exist. 7Nevertheless, so that nothing is omitted, a vowel then follows so each name will consist of 4 lettersintotal...I should perhaps continue with examples, but as I am neither a botanistnoranapothecary,Icannotsupporttogofurtheroutside myfield(Polhem1954,350-351).
Polhem's successors seemed to agree.His ideas on this matter were neglected.
NathanielMatthaeusvonWolf'sScheme (1776,1782) The next person to have published on this topic was Nathaniel Matthaeus vonWolf(1724-1784).8VonWolfwasborninKonitz,atowninwesternPrussia, onJanuary24,of1724,thesonofanapothecary.9 Heearnedhismedicaldegree in 1748 in Erfurt and built up a good reputation as a physician and as an astronomer.In 1766, he was elected to the Polish nobility.Von Wolf visited England from November 1759 to July 1760 in order to pass the six months on EnglishsoilrequiredatthattimetoallowhimtoqualifyasaFellowoftheRoyal SocietyofLondon(RoyalSocietyofLondon2006).HewaselectedasaFellowon April 10, 1777.In 1776, von Wolf came out with Genera plantarum vocabulis characteristicis definita, a Latin publication expanding on ideas of plant classificationhehadbeenworkingonforseveralyears.
substance, superficial qualities, location, shape, smell and flavour.Each manifestationofthesenotionswastoberepresentedbyaletter.Forinstance, for the different kinds of substance that plant parts could manifest, von Wolf assignedthefollowingletters: Substancetype Letter Theensembleofhistechniquewaslikewisedisplayedasachart(Figure1).Instead of using uninomials, as did the other inventors of similar schemes, von Wolf split his plant names into two parts so that they were a kind of binomial.One name stood for the plant's family and the other for the genus, though in other respects the generation of the names was the same.He suggestedthatplantfamilynamesshouldbetwoletterslong,thefirstindicating the number of pistils (or pistilliform stamens), the second, the number of stamens.Adding a prefix "h" would indicate an unequal number of stamens, whileadding"z"wouldindicateanabsenceofstamens.Forgenericnames,the firstsyllabledescribedthefruit.Thenextsyllablewastodescribethenumberof involucres,thenextthecorolla,andso on.Heprovidestheexampleofaplant withonenakedseed:accordingtohissystem,thebotanistwouldfirstwrite"A" and"ń."Noticingthatthisparticularflowerhasfivesimplepetalsinatube,hewouldadd"f"and"i"tomakethegenericname,Ańfi.Theflowerhasonepistil and five stamens, making the next syllable, or family name, Aî.The botanist couldthenlookupamoredetaileddescriptionfortheplantunderitsfamilyand genus(AîAńfi),wherehewouldalsofinditsLatingenericname,Calligonum(von Wolf1776,7).Figure1.N.M.vonWolf'schartexplainingwhichlettersstandforwhichplantproperties.
VonWolfusedhistechniquetobrieflydescribeandnamenew"families"and to provide new "generic" names for hundreds of plants.Still, his dedication to the botanical cause seems ambiguous.The Genera plantarum was initially published anonymously, and on very bad quality paper.The index-published separatelyin1780-isamarvelofdisorder,lackingalphabetizationbeyondthe firstletterandmissingmanyentries.Itshowsataglancethatsomeplantshave multiplenamesandotherplantsshareonename.VonWolfneverthelessmadea gift of his book to the Royal Society on November 9, 1780. 10The Abbé des Houssayes (1727-1783) also, quite astonishingly, did not mention anything unusual about the index in his scathing review of von Wolf's work in 1781though he called von Wolf's nomenclature "hieroglyphics," "barbarous," and "unintelligible"(Deshoussayes1781,405).
Despite these drawbacks, von Wolf's work was popular enough to merit a second, expanded edition published in 1782 and titled Genera et species 10 PresentsMadetotheRoyalSociety,1781."N.M.deWolff,M.D.,F.R.S"alsopersonallygave an octavo copy of the second edition of his work to the Royal Society on November 13, 1783 (PresentsMadetotheRoyalSociety,1784).
Doctor Jonathan Stokes, compiler of references to figures in the English botanist William Withering's very popular Botanical Arrangement of British Plants( 1787),wasfamiliarwithvonWolf's1776edition.Of"Wulff,"hewrotein anasidethathehad onceformedabotanicallanguageonaplansomewhatsimilar,but I soon discovered that in proportion as plants resembled each other, the difficulty of distinguishing the sounds or combinations of letters expressive of them, must proportionately increase.Languages formed on plans of this kind, must be full of such ambiguous names as Clutia, and Clusia (Withering and Stokes, 1787,xlvi).
(Weiss 1854).He took courses in surgery, anatomy and natural history at Bordeaux, then moved to Paris.Benoît Dayrat indicates that he was born in 1752, in Oléron, in the Béarn district, slightly south of Lasseube, though his parentsmovedtoPariswhenhewasyoung.Aftertheydied,hedecidednotto return to his birthplace with his relatives but instead to stay in Paris to study surgery and take courses in botany with Bernard de Jussieu.He was never enrolled in a school or faculty of medicine (Dayrat 2003, 160).Either way, in 1776heundertookadescriptionoftheplantsoftheParisregion,butsetaside thisworktoprepareandthenteachabotanicalcoursehestarted.
It was during this time that he started to compile the material that would grow into his masterpiece, "Phytonomatotechnie universelle [a universal techniquefornamingplants],or,theartofgivingplantsnamestakenfromtheir characters; a new system by the means of which one can, by oneself, without thehelpofanybook,namealltheplantsthatgrowonthesurfaceoftheearth" (Bergeret1783,157).Bergerettestedhissystemonhisstudentsbeforehavingit published."Ondifferentfieldtrips,"hewroteintheprospectustohisbook,having put the principles of my system into their [i.e.students'] hands,Ihadthesweetsatisfactionofseeingthemgraspit,andI saw with pleasure that, by the means of these very principles, they were brought without trouble to name to me phytonomatotechnically all the plants that they found at their feet.The displays of joy on their part were not at all equivocal; theysawthemselves,sotospeak,asthecreatorsofnames,and this joy became even more evident when they recognized the appropriateness of the application of the letters to the different characters,accordingtotheconformityofthesenameswiththose thatIhadalreadylaiddown(Bergeret1783,157-158).

Allthesechartsareputtogetherinamannersuchthateachone
gives a letter to the plant, according to its characters [i.e. the plant's features]; the resulting name can be spelled easily (Bergeret1783-1784,i[unnumbered]).
Eachnamewastobe15letterslong.Unlikealltheotherschemesdiscussed here, his did not involve strict alternation of vowels and consonants.Names containing several of the same letters in a row-such as those in which many parts were assigned "A" for "absent"-could be condensed by adding a superscript number over the first repeating letter.This was to indicate how many of the same letters were to follow.For instance, the phytonomatotechnical namefortheflyagaricmushroom,AAAAAAAALAAAAYZ, could be spelled A 8 LA 4 YZ for short (Bergeret 1783(Bergeret -1784, i [not numbered], i [not numbered]).He then ran through how to generate these names, using the example of Belladonnaofficinalis(deadlynightshade): Wecomparethecorollawiththecharactersofthefirsttable;we seethattheletterJindicatesacorollainfivesections.WewriteJ andmoveontothesecondtable,whichgivesthevowelEbecause the sections of the corolla are not cut very deep.We put the E nexttotheJ,andsowehaveJE.Wegotothethirdtable,andfind thattheletterQindicatestheinsertionofthestamensunderthe germ, by means of the corolla.We write Q, and have JEQ.The numberofstamensisfive,sowefindinthefourthtablethatthe letter L indicates those for this plant.We write it and we have JEQL.ThesideopeningoftheanthersisexpressedbytheletterY inthefifthtable;wewriteitdownandhaveJEQLY.Finally,bythis method,wetravelthroughthefivetablesinturn,eachfurnishing a letter; we write them down, and obtain the name JEQLYABIAJISBEV, which is equivalent to the entire description that Linnaeus gives us of the genus of deadly nightshade, since eachoftheselettersexpresses,aswewillsee,acharacterofthe plant (Bergeret1783-1784).
today, copies of Phytonomatotechnie are valued more for the high production standardsoftheplatesthanforBergeret'ssystematictreatmentofplants.
He seems to have lived in Arras and may have also taught chemistry there duringthe1780s(Dalmasso2005.)In1783,hepublishedhisbotanicalschemeinaworkcalledPhytographieuniverselle,ousystêmedebotaniquefondésurune méthode descriptive de toutes les parties de la fleur: avec une nouvelle langue antho-phyllographique.It does not seem to have attracted much notice other than Palisot de Beauvois's dismissal and scorn.But de Las himself was quite proudofhisworkandconvincedofitsimportance.Ofallthedud-makersIhave found, he is the only one who acknowledged the existence of other authors' similar schemes.In particular, he wrote off Bergeret's Phytonomatotechnie as inadequate.Healso insistedthathehadnotplagiarizedfromBergeretorfrom vonWolf.Clearly,hesawBergeretandvonWolfashisrivalsinthedevelopment ofthebestvariationofanentirelyfeasibleproject.
I know quite well that botany does not consist only of the knowledgeof[namesof]plants,butIalsoknowthatitisthemost necessarypartforbeginners..[but]memoryrefusestokeeptrack of an infinity of words that have no relationships among themselvesnorwiththethingstheysignify.Ifanalgebraicformula does not bring up the principles that served to build it, do you thinkIcanretainit?Theconnectionithaswithitsprinciplesfixes itinmymind.
Notonlywilleverymaninstructedinthecharactersofflowersas my method indicates be able to recognize the plant that each anthographic word signifies without worrying about making mistakes, he will be equally capable of composing the name of eachplantthathewillhavebeforehiseyes,withthecertitudeof making it known to all people of whatever nation they are who arefamiliarwithmycharactersandthesignsthattheyrepresent (deLas1783,50-54).
His anthographic rules for naming genera were sufficiently complex that it took de Las 17 pages of text to explain all of their principles, corollaries and exceptions.In brief, they involved a modified Greek alphabet in which each letterstoodforacharacteroftheplantandhadplacevalue.Vowelsoundswere to be added between the consonants at will to make the name sound harmonious(seeFigure3).Hisphyllographiccharacters,orspecificnames,each consisted of two consonants, standing for two outstanding features of the plant'sleaves.Again,vowelsweretobeaddedtotheseconsonantssothatthe resultingtwo-syllablenamewouldbeeasytopronounce.Heincludeda9-page "Phyllographic alphabet" indicating which letters stood for which features andwhen,supplementedbya255-word"Phyllographicdictionary"definingeach ofthewordsdescribingthefeaturessoastoremoveallambiguity.
TheAbbéwasawarethathisproposalwasnotmainstream.Infact,hewrote that "fear of the censure of intellectuals on this matter would have made me holdbackfromprinting[it],ifM.Bergerethadnotpublishedtheprospectusof hisPhytonomatotechnie,whichmyworkresemblestoomuch,toputmeatrisk ofbeingreproachedforplagiarismifIweretodeferpublicationanylonger"(de Las 1783, iv).Bergeret's prospectus was published in May of 1783 ("Phytonomatotechnie Universelle," 1783).De Las quickly went on the defensive,preparingtopublishassoonaspossible.Heensuredthathisreaders knew that he had given copies of his project to the leading botanists Michel Adanson and Antoine-Laurent de Jussieu in November 1782, and that Jussieu had given him suggestions on how to improve it.He listed a number of other men of letters with whom he had communicated about his ideas, including severalfromArrasandothersfromDijonandLyon."Almostall"ofthesemen,he wrote,"havedonemethehonourofwritingmethattheywouldpermitmeto calltheminproofoftheknowledgethattheyhadofmywork,longbeforeMr.BergeretpublishedtheprospectusofhisPhytonomatotechnie."Hereprintedthe attestationofMarcAntoineLouisClaretdeFleurieudelaTourette(1729-1793), perpetual secretary of the Académie de Lyon, "against whom I doubt that M. Bergeret would dare to bother to argue with me over [s'élever pour me contester] the glory of having conceived at the same time as him a project favourabletotheprogressofbotany"(deLas1783,v).
Aswell,deLasmadesuretoexplainthathehaddevisedhisschemebefore he had ever heard of "Mr.Wolf, the Polish physician."Adanson, de Las wrote, firstmadehimawareofvonWolf'swork only in conversation.M. de Jussieu showed me the book, which did not rest in my hands, as the same M. Jussieu can attest, for more than around four minutes.However brief this time was, I believeit ispossibletoassurethat itsufficedformetoseehow barbaroushismethod is.Itismoreofawriting thanalanguage.How to pronounce, in effect, a language that is made up of twenty-five vowels, five As, five Es, etc., each of these As and these Es differing only in the pronunciation taken from its neighbours, of the sort that to speak the language of Mr. Wolf, onemustknownotonlyGerman,butalsoallofitsdialects.What confusion!Couldtheauthoreverpromisethathislanguagewould beadopted?DeLaswasemphaticthathisownmethodwaseasiertopronouncethanany other.He demonstrated that, for instance, according to von Wolf's plan, the letterB signifiesthatthefruitisaberry,Âwitha circumflexaccent,that theflowerhasfourpetals,W,thatthecalyxisconical,andplainE that the calyx has four segments.[The Linnean genus] Actaea, placed by this author in the second class, distinguished by all its characters,isnamed byMr.WolfBåê We,whichisequivalent in pronunciation to Baaeeoue, and consequently cannot be pronounced.The same plant is namedCucagexpi in my method, and is more completely described, because this word includes again the form that the petals take among themselves and the numberofstamensandthepistil(deLas1783,58).
Combining the Abbé de Las's anthographic genus names and phyllographic species names makes for a particularly awkward nomenclatural scheme, one thatwasneveradopted.
The final dud-maker whose work I located was, like John Wilkins, an Englishman.Nevertheless, his ideas about botanical classification were clearly developed independently from Wilkins' and all other preceding schemes discussedabove.
HobartTown,thecapitalofVanDiemen'sland,wasonly24yearsoldwhen Hendersonarrived.Itspopulationwaslargelyroughanduncultured,asbefitted this former penal colony in such an isolated location.Still, by January 1830, Henderson had persuaded "most of the respectable settlers throughout the Island" to form the Van Diemen's Land Society, otherwise known as the Philosophical Society or the Van Diemen's Land Scientific Society (Hoare 1968, 12).
Henderson,asPresidentoftheSociety,gavealongspeechbeforemorethan 100 of the most influential islanders, who gathered in the courthouse for the firstmeeting.TheHobartTownCourierreportedonJanuary23,1830,thatatthe heartofHenderson'stalkwasaproposalfor an entirely new system for introducing one general and determinate form of expression by which those who collected new plants, animals and other curiosities, through at a distance from each other, might infallibly be enabled to give the same name to their discoveries… [He had waged] war against 30,000 arbitrarynamesofplantsreceivedinthenomenclatureofbotany, andhadsuggestedthesubstitutionofcertainsyllablesandletters, of which might be compounded names expressive of the diagnosticmarksofeachparticularplant(Hoare1968,14-15).
Henderson later claimed to have been inspired to develop these ideas not from experience with botany, but from observing that judges presented with practicallythesamelegalargumentswerelikelytointerpretthemdifferently.He surmised that a logical outline of how to proceed in each instance would help reduceuncertaintyandvariationineachlegaldecision.Ifsuchasystemwerein place, "legal language would . . .become brief and determinate; each legal document, henceforth constituting of a series of definitions, arranged in a regular and successive order," would direct judges to follow the same, proper path(Henderson1832,176-177).Hendersonthoughtthatbotanywasadriftina seaofunintentionalsynonymycausedbyalackofguidelinesonhowexactlyto name plants.He believed it could benefit from this kind of logical overhaul as well. 15 The talk was "received politely and well," though "opinion was divided" about it.A prominent physician argued in favour of naming local plants according to Henderson's principles, while the publisher of the Hobart Town Courier thought that classifications developed in Europe by expert botanists werebetter.HerecommendedtheLinnaeansexualsystem.Thereisnorecordof anyone in attendance mentioning Robert Brown's well-respected Prodromus floraeNovaeHollandiaeetInsulaeVan-Diemen (IntroductiontothefloraofNew Holland[Australia]andVanDiemen'sIsland),aLatinworkpublishedinEngland in1810andarrangedaccordingtoJussieu's"naturalmethod"(Hoare1968,15).15 Henderson explainedhisideasindetail ina letter,describedbelow.Healsodevotedseveral pagestoexplaininghowdividingplantsintosuccessivepairsofgroupsofequalsizewouldbethe ideal waytoorganize informationaboutthem.Hisfixationondichotomousdivisionsmayhave beeninspiredbysimilarproposalsputforthinJeremyBentham'sChrestomathia(1817),orJohn Fleming's Philosophy of Zoology (1822), with which he shares the unusual terminology of "positive and negative" divisions.(His apparent inability to read French makes familiarity with George Bentham's more extensive and more botanical Essai sur la nomenclature et la classificationof1823unlikely).Henderson'sproposalalsoresemblesGottfriedWilhelmLeibniz's 18 th -centuryideaforauniversallanguagedesignedtopreventillogicallegaldecisionsfrombeing made,whichwas,inturn,inspiredbyBishopJohnWilkins'Essaytowardsarealcharacter(1668) (Maat2004,156,301-302).Henderson'sproposal,however,wasnotliftedwholesalefromanyof theseworks.Italsocontainsenoughoriginalelements-includingflawsaddressedinsomeofhis predecessors'schemes-tosuggestthathedeviseditsdetailsonhisownevenifhehadheardof similar ideas from others.In several ways, Henderson's proposal may be the botanical nomenclature analogue of the Cherokee writing system, a system devised by the illiterate, unilingualSequoyahafterhesawEnglish-speakerscommunicatingbymeansofwrittensymbols.SeveralotherNorthAmericanIndianwritingsystemsweresimilarlyinventedinthisway (Walker 1981,145-147,151-152).
This English-language treatise speaks well to what those familiar with Henderson's dealings in Hobart Town characterized as his "censorious and dogmatic" character (Hoare 1968, 19).He began by stating bluntly that he preferred to send his ideas to France, the home of Lavoisier, reformer of chemicalnomenclature,thantotheRoyalSocietyinEngland.Hedidnotexpecta "favourable reception" from the English, who, he wrote, had "firmly rooted prejudices" "inimical" to nomenclature reform (Henderson 1832, 156).He followedtheseremarkswithabloatedandvituperativelistofstandardbotanical complaints:therearetoomanykindsofplantstoremember;toomanyabsurd andarbitraryplantnames;toomanysynonymsanddisputednames;toomany technical terms; and too many botanical works to look through in order to identifyplants.Henderson proposedthatastudyofplant physiologywaswhat really mattered, and it would be best served if botanists abandoned artificial systemswrittenindeadlanguages(Henderson1832,157-159).
AssumingthatQuatermèredeQuincyhadcontinuedreadingpastthisgaffe, hewouldhavecometothecruxofHenderson'sargument.Hendersonwishedto propose"theformationofaregularsystemofnomenclature"suchthatitwould beclearlyworking when an individual, who might in one country discover a new chemicalsubstance,mineral,plant,insect,orotheranimal,should be enabled by means of the system to give it the self-same appellationthatanotherindividual,havingnocommunicationwith the first, would have assigned to it, had he discovered its copartnerinanyotherpartoftheworld [Henderson'sitalics].
He wrote that he intended "to prove, not only that such a system was perfectly practicable, but that it could be likewise reduced to simple and easy regulations"-unliketheLinnaeansexualsystem,whichhethencriticizedagain atlength(Henderson1832,160).
Henderson'sexplanationforwhyhisproposedsystemwouldbeeasiertouse thanthesexualsysteminvolvedanoutlineofhowhebelievedhumanmemory towork.Heacknowledgedthatoralhistoriespassedonaspoemsare"lesseasily corrupted, and more easily recollected than prose," attributing this to their regular structure of a narrative "chain of ideas" reinforced through rhyme.Linnean nomenclature, however, offers "neither line of connection betwixt the specimen and the name, or betwixt one name and another."When names are givenwilly-nillylikethis,"themindendeavourstosupplythedefectbyanideal chain of its own construction."In general terms, an "unharmonious" order of materials presented to a learner causes difficulties in recollection (Henderson 1832,162-163).
Henderson consequently proposed "to simplify and establish a general systematic nomenclature in science" by "rendering as determinate as possible our methods of arrangement" such that each name would "describe . . . the exactpositionofthespecimenintheestablishedsystemofclassification."Sincehewishedhisclassificationschemetobeasnaturalaspossible,theclassification andnomenclatureforeachplantneededtobebasedon"somefactorqualityin the class or specimen" (Henderson 1832, 168, 175).Henderson considered it feasible for each class, order and genus of plants to be defined according to a propertysharedamongallitsmembers,somethingbotanistsmorefamiliarwith botanical classification knew was impossible.Numerous species of plants have different numbers of stamens and/or pistils from other plants considered to belongtothesamenaturalgenera.Oneexamplediscussedatlengthinbotanical circlessincethe1760sisthespeciesValerianarubra.Ithasonestameninstead of the three found in other species of valerian (Müller-Wille and Scharf submitted2008).
Henderson's next steps were to designate each class with a different consonant and to assign a different vowel to each order.The genera in each order were likewise each to be represented by a consonant, while the species weretobelabelledwithvowels,"resortingtotheemploymentofdoubleletters, diphthongs,oreventoconsonants,whereverthehighernumbersmaybecalled for" (Henderson 1832, 176).This method of combining alternating consonants and vowels, each with its own descriptive significance, would generate meaningfulandpronounceablenamesoforganisms.AsHendersonexplained, the name 'Bal' will clearly represent the first subdivision of Monandria Monogynia [plants with one stamen and one pistil]; and Balba will equally represent the first species, in the first genus,inthefirstsubdivision,inthefirstorderofthefirstclassof the Linnaean system.One point more yet requires to be determined;namely,towhatsciencethisnamemaybelong;and wemustthereforeappendtoitsometermination,suchas'na,'or Balbana,inordertoindicateitsbeingthenameofaplant,andto distinguish it from that of an animal or mineral.The name 'Dombina'willinthesamemannerrepresentthethirdspeciesof the second genus, in the second subdivision of Tetrandria Tetragynia [plants with four stamens and four pistils]; the termination 'na' becoming the symbol of the class-word, indicatingthereby,itsrelationtothedepartmentofBotany.
Any living thing or mineral could easily be named in this way, achieving Henderson's goal of a system of nomenclature that would enable anyone, anywhere,tospontaneouslycomeupwiththesamenameforthesamekindof organismormineral (Henderson1832,176-178).
ThereisnorecordofareplytoHendersonfromtheInstitutdeFrance.Inall likelihood, Quatermère de Quincy ignored the missive as yet another waste of paperfromanuneducatedcrank.
As Kuhn (1959, 344 note 348) and Troyer (1992, 838) note, using vague criteriaaboutwhatcountsasamultipleinventionwillmakedissimilarinventions seemalike,whereasusingoverlystrictcriteriawillmakesimilarinventionsseem dissimilar.Whilethestrictnessofcriteriaaresubjective,Ithinkitisreasonableto attempt to take a middle road, as Misa (1994) suggests, and assume that the dud-makers were working on the same sort of project, despite the individual differences.The differences can then be taken as evidence in support of the independence of the inventors.De Las's opinions of Bergeret's and von Wolf's schemes as fundamentally similar in aim though not in execution substantiate this.Table1summarizesthedifferencesamongtheschemes.
Asisevidentfromthedatainthefollowingtable(Table1),therearenotwo inventors whose schemes look exactly alike.The languages they spoke, the languages in which they wrote their proposals, and the precise details of the formsofthenamesthattheirschemesweremeanttogeneratearevariable.I have found only one obvious trend in their output: that is, the influence of Linnaeus on the characters acceptable to use for plant identification.In his Genera plantarum (1737), and in subsequent publications, Linnaeus restricted the kinds of evidence acceptable to use in descriptions of plants to number, figure, position, and proportion, banning reliance on colour, smell, taste, and other features dependent upon a plant's effect upon the person describing it (Linnaeus 2003, 219-236;Müller-Wille 2005, 88).These rules became widely accepted among botanists.Bergeret, de Las and Henderson obeyed these conventions,incontrasttoKinner,Wilkins,PolhemandvonWolf.Thischangein theconventionsofwhichcharactersaresuitabletouse in plantdescriptions is indicative of a basic shared awareness of at least some contemporary ideas in botany.As well, Kinner, von Wolf, Bergeret, and Henderson were physicians.This automatically gave them further familiarity with plants.The dud-makers were also all intellectually adventurous and motivated to succeed in their respective professions.Further, even the men who were not physicians received formal educations.They may therefore have all been exposed to similar perspectives aboutthefunctionsof languageandmemory, particularlyscholasticideassuch asLlulliancombinatoricsandsyllogisticlogic.
Leibniz was certainly inspired by Llullist and Ramist conceptions of knowledge, especially through his readings of the philosophical language projectors George Dalgarno (1626?-1687, another member of Hartlib's circle) andWilkins (Maat2004,69,184,270;butseealsoLewis2007-132).Hetookthe ideaofcombiningunitattributesinmanydifferentwaystoanewlevel.Henot only envisaged the generation of new ideas by combinatorial means, by transporting the permutation and combination techniques he developed in mathematicstotherealmofhumanthoughtingeneral;healsowrotethatifall ideascouldbereducedtotheirbasiccomponents,andifthosebasicelementsof thoughtcouldberecombinedsystematically,allknownideasaswellasnewand interestingideascouldbegenerated(Maat2004,273).Leibniz'semphasisonasof-yet-undiscovered ideas becoming obvious because of the possibility of exhausting all logical combinations was unusual; most of his predecessors and contemporariesdidnotquantifythenumber ofideasthatwouldbegenerated ormakerecoursetologicinthisway.Mostimportantly,Leibnizalsoperceived thatmanyoftheideasgeneratedthiswaywouldbeirrelevant.Hesawthattheir number would also be so vast that nobody would have the time to deal with themallonebyone.Hesuggestedthata"methodofexclusions"wouldbeaway toseparatethewheatfromthechaff(Davies1986,267).Thismentalwinnowing technique, like many of his other ideas concerning the complete enumeration and systematization of everything and the perfection of human thought, was moreofadesideratumthanafunctionalalgorithm.Neitherhenoranyoneelse wassuccessfulindevelopinganapriorimethodtoseparatetheusefulfromthe fanciful or the existent from the possible.Despite the centuries of efforts put into them, scholarly attention to combinatorial idea-generation techniques began to fade by the 18 th century (Rossi 2000, xviii, 44).The evolutionary epistemology of trial and error-the only way that worked to pick out good ideas-wassimplynotquickorlogicalenough(Campbell1974).
Combinatorialdescriptions,however,didexperienceacertainformofrevival in a discipline in which they still promised some utility-botany (Rossi 2000, 171).The technique survived in botany not because it had any absolute predictivevalue,butratherbecauseitprovidedatoolfordescribingplantsthat workedwellwiththeknowledgeofplantsthatwasavailableinthe18 th and19 th centuries,asdescribedabove.

MnemonicNamesforSyllogisticMoods
Though details of their formative years are few, the dud-makers likely also sharedexposuretothesamemnemonictechniqueusedinphilosophyclasses.It wasnormaluntilthelogicreformsofthe19 th centuryforthemedievalnamesof categories of syllogisms, called moods, to be explained in university courses throughtheuseof24odd-soundingmnemonicnames: Barbara,Celarent,Darii, Ferio,Barbari,Celaront,Cesare,Camestres,Festino,Baroco,Cesaro,Camestrop, Darapti, Disamis, Datisi, Felapton, Bocardo, Ferison, Bramantip, Camenes, Dimaris,Fesapo,FresisonandCamenop.Eachofthethreevowelsinthenameof amoodindicatesinturnthecategoricalformsofthetwodifferentpremisesand the conclusion of all syllogisms that are of that mood.The consonants yield information about how to convert syllogisms of the last twenty moods into syllogisms of the first four moods through a series of logical operations somewhat akin to proving trigonometric identities. 16Youthful familiarity with these coded, descriptive, meaningful and pronounceable names may have implanted the idea in the dud-makers' minds that plant names formed in a similar way could solve the problem of synonymy.However, none of the dudmakers referred to the naming of syllogistic moods in his works, and none claimedtohavemadesystemsthatwereintended,assyllogisticmoodswere,to exhaustalllogicalpossibilities.Chancesarethatwhateverinfluenceknowledge ofthesyllogisticmoodnameshadonthedud-makerswasnotintheforefrontof theirmindswhentheywereconstructingtheirschemes.
Considering that the men grew up in different economic and cultural circumstancesandalsohaddifferentinterests,theremustalsobeotherfactors at play to the dud-makers from their contemporaries who shared similaruniversityeducationsandaninterestinbotany,butwhodidnotgoonto create duds.Whatever these factors were in the lives of the individual dudmakers, however, is likely to be impossible to determine, given both the fragmentaryrecordsoftheirlivesthathavebeenpreservedandthedifficultyof pinpointingexactsourcesofscientificinspiration.
Botanists with better access to collections and greater networks of collaborators than the men whose schemes I described realized that schemes basedonmemorizationalonewerenotgoingtowork.Inthe17 th ,18 th ,andearly 19 th centuries,speciesofplantswerethoughttobethemostnumerouskindof thingintheworld. 17Asmentionedintheintroductionabove,CasparBauhinwas saidtohavedescribedover6,000plantkindshadinhisPinaxtheatribotaniciof 1623.Authorsofmorewidely-acceptedbotanicalschemesfrequentlymentioned thenumberofentriesinthePinaxwhendiscussinghownumerousplantkinds are.They were aware that whatever schemes they were to develop had to distinguish these memory-busting numbers of kinds.And although each of the proponents of meaningful plant names was concerned with the pronounceability,regularityandbrevityofthenamestheydesigned,thenames themselveswerefarfrommemorable.Kinner,Polhem,vonWolfandtheothers had all made the same mistakes in regarding human memory as operationally equivalenttothe"artificialmemory"ofthewrittenword.17 Whetherthe"species"holdupasvalidspeciestodayisbesidethepoint,sincewhatmatters hereisthenumberofkindsofthingsthatneededtobedistinguishedintheschemes.
For instance, while Henderson was correct that memory is helped by narrative, rhyme, and order of a sort, he and all of the other scheme-builders describedhereneglectedtotakeintoaccountthatimagery,spatiality,dramatic elements, metaphor and metonymy are also effective mnemonic aids.The centuries-old tradition of ars memoria, based on these techniques, may have been dying in the 17 th century, but it was still in use among lettered and unlettered folk alike during the time when at least Kinner and Wilkins were writing(Rossi2000,originalItalianpublication1975).Nodud-makerdescribedin this paper, however, refers to or seems to make use of this tradition.Instead, whattheirnamesmadeupforinregularity,theylostinsalience.Essentiallywhat thesemenhaddonewastotaketechniquesthatworkedwellinprint-afixed order, coded short forms, tabular layouts-and assumed that they would transferwelltomemoryandspeech.Theresults"speak"forthemselves.
Thiskindofthinkingaboutthought,Ishouldadd,wasnotconfinedtodudmakers.A number of their contemporaries also seemed so comfortable with books and their layouts that they imposed these arrangements on their conceptions of how the natural world is structured.When the English botanist JohnHill,forinstance,proposedin1759thefoundationofabotanicalgardenfor publicinstruction,hewrotethatsuchagardenwouldbe"akindoflivingherbal" (Hill 1759, 17).Hill's remark shows that the printed page as a means of organizing natural history was well entrenched in enlightenment culture.Hill broughthiscomfortwithbookswithhimintothefieldandturnedherbals,which weremeantasstand-insforlivingplants,intomodelsforarranginggardens.All of these examples suggest that we should be extra-careful not to justify classifications according to the techniques or metaphors that happen to be trendy at a given time.As we push the envelope of each new tool for investigating and maintaining information about nature, we also confine ourselvestothinkingaboutnatureintermsofourtools'limitations.
A friend of mine, systematic botanist Tim Dickinson, explained to me years ago that any good system of botanical organization has to have local memorizability,sothatbotanistscanrecallthenamesofplantsthattheywork with, and global "look-upability," so that unfamiliar plants can be identified quickly.The sheer number of plant kinds necessitates awritten record for this purpose.And if a record is written, it is best recorded in terms that a casual readercanunderstand:thatis,inordinarywords,whichdonotrequireusersto memorizebothtechnicaldistinctionsandtheirlettercodes.Kinner,Polhem,and the others did not work hands-on with enough plants when they were writing theirschemestohaveencounteredtheselimitsofhumanmemory.

Aftermath
Although some isolated people were still promoting combinatorial plant names into the 19 th century, the majority of botanists and even laypeople by thattimewerecynicalenoughtogreetproposalsofthistypewithsarcasm.For instance, a notice in the Athenaeum, a popular magazine, stated, "We understand that some German botanists are labouring at the invention of cabalisticcharactersforplants…wesupposetheAlgebraicbotanywillbecalled anaturalsystem!!" ("AlgebraicBotany"1828)Thetwoexclamationmarksleave no doubt about how many combinatorial schemes and so-called "natural systems"thepublichadseencomeandgobefore.
Curiously, one might say, dudsof the kind described earlier did not remain dudsforever.Rustyandabandonedtechnologiescanberefinedandtransported to new contexts where they will shine.Removing the need for coded plant descriptionstobenames,forinstance,resultsinakindofshorthand.Anumber of 18 th century botanists had toyed with symbolic ways to describe plant features to compress information in this way, including Jean-Jacques Rousseau (Cook 2004, 80-87).Peter Stevens has described a number of other formalized floral diagrams and floral formulas from the early 19 th century (Stevens 1994, 139-141;StevensreproducesapageoffloralformulasfromSeringeandGuillard 1836).Like shorthand, floral diagrams and formulas can be useful tools for describing plant structure in a compact and efficient way.And like shorthand, theyarenotmeanttobereadaloudasanythingbutaseriesofwords.
Maintainingthereferencefunctionofdescriptivecodednamesandtheplace valuesofeachletter,butremovingtheneedforthemtobepronounced,yieldsa way to make call numbers.Schulte-Albert's work on Kinner praises him for developing a "faceted classification" much like those now used in library catalogues. 18Nobodymemorizesthousandsofcallnumbers,buttheyareuseful written reminders of where books are located.Modifying the schemes in another way produces "hash tables."Theseare essentially one-line, computergenerated summaries of database entries, also used as their addresses.And Henderson's logical guidelines for deciding legal cases and for identifying and naming plants can be easily turned into flow charts or computerized "expert systems," used for similar diagnostic purposes since the 1960s.Evidently, substitutingwritten"memory"orcomputermemoryfor humanmemorycould givethesedudsnewlife.
This situation should raise a historiographical flag regarding the extent to which scientific creativity may be technologically determined, or, at least, constrained.Cases of multiple duds becoming multiple inventions support the ideathat,toparaphraseGregRadick'sassertionabouttheories,inprinciple,any unsuccessful invention may have the potential to succeed with the addition of therighttechnologicalfix(Radick2006,31).Ifdudscanbeconvertedtosuccess stories merely by the addition of new technologies, we need to look in more detailattechnologiesthatwererejectedintheirowntimes.Onlythenwillwebe 18 Infact,thecolonclassification"facets"ofS.Ranganathan,themannowhailedasthefatherof faceted library classifications -personality, matter, energy, space and time -bear an uncanny resemblance to Aristotle's predicaments or categories -substance, quantity, quality, relation, place,time,position,state,action,andpassion-asHansSchulte-Albertpointedout (1974,324).
able to give a more complete account of what processes give rise both to the contingencyandtheinevitabilityofbothdudandsuccessfulmultipleinventions.