耶鲁大学著名生态学与进化生物学教授写给研究生的一些忠告

耶鲁大学著名生态学与进化生物学教授写给研究生的一些忠告Stearns教授的一些犀利的观点，你认同吗？------------------------------------------------------------准备坏的结局。没人（包括有些教授）会在乎你；主动接近导师。应该明确为什么你的研究重要。心理问题会成为最大障碍。不要选修课程：太低效。找人批评你的研究计划。“管理”你的导师：让他知道你的存在。在不同的论文形式中选择适合你的挑战方式。早发文章（署名、合作等有技巧）。不用小看硕士论文。经常性发文章，但别太多。@Zurich------------------------------------SomeModestAdviceforGraduateStudentsbyStephenC.StearnsAlwaysPreparefortheWorst.Someofthegreatestcatastrophesingraduateeducationcouldhavebeenavoidedbyalittleintelligentforesight.Becynical.Assumethatyourproposedresearchmightnotwork,andthatoneofyourfacultyadvisersmightbecomeunsupportive-orevenhostile.Planforalternatives.Nobodycaresaboutyou.Infact,someprofessorscareaboutyouandsomedon't.Mostprobablydo,butallarebusy,whichmeansinpracticetheycannotcareaboutyoubecausetheydon'thavethetime.Youareonyourown,andyouhadbettergetusedtoit.Thishasalotofimplications.Herearetwoimportantones:1.Youhadbetterdecideearlyonthatyouareinchargeofyourprogram.Thedegreeyougetisyourstocreate.Yourmajorprofessorcanadviseyouandprotectyoutoacertainextentfrombureaucraticandfinancialdemons,butheshouldnottellyouwhattodo.Thatisuptoyou.Ifyouneedadvice,askforit:that'shisjob.2.Ifyouwanttopicksomebody'sbrains,you'llhavetogotohimorher,becausetheywon'tbecomingtoyou.\nYouMustKnowWhyYourWorkisImportant.Whenyoufirstarrive,readandthinkwidelyandexhaustivelyforayear.Assumethateverythingyoureadisbullshituntiltheauthormanagestoconvinceyouthatitisn't.Ifyoudonotunderstandsomething,don'tfeelbad-it'snotyourfault,it'stheauthor's.Hedidn'twriteclearlyenough.Ifsomeauthorityfiguretellsyouthatyouaren'taccomplishinganythingbecauseyouaren'ttakingcoursesandyouaren'tgatheringdata,tellhimwhatyou'reupto.Ifhepersists,tellhimtobugoff,becauseyouknowwhatyou'redoing,dammit.Thisisahardstagetogetthroughbecauseyouwillfeelguiltyaboutnotgettinggoingonyourownresearch.Youwillcontinuallybeaskingyourself,"WhatamIdoinghere?"Bepatient.Thisstageiscriticaltoyourpersonaldevelopmentandtomaintainingtheflowofnewideasintoscience.Hereyoudecidewhatconstitutesanimportantproblem.Youmustarriveatthisdecisionindependentlyfortworeasons.First,ifsomeonehandsyouaproblem,youwon'tfeelthatitisyours,youwon'thavethatpossessivenessthatmakesyouwanttoworkonit,defendit,fightforit,andmakeitcomeoutbeautifully.Secondly,yourPhDworkwillshapeyourfuture.Itisyourchoiceofafieldinwhichtocarryoutalife'swork.Itisalsoimportanttothedynamicofsciencethatyourentrybewellthoughtout.Thisisonepointwhereyoucanstartawholenewareaofresearch.Remember,whatsensedoesitmaketostartgatheringdataifyoudon'tknow-andImeanreallyknow-whyyou'redoingit?PsychologicalProblemsaretheBiggestBarrier.Youmustestablishafirmpsychologicalstanceearlyinyourgraduatecareertokeepfrombeingbuffetedbythemanydemandsthatwillbemadeonyourtime.Ifyoudon'twatchout,thepressuresofcoursework,teaching,languagerequirementsandwhoknowswhatelsewillpushyouaroundlikealarge,docilemoleculeinBrownianmotion.Hereareafewthingstowatchoutfor:1.Theinitiation-ritenatureofthePhDanditspowertoconvinceyouthatyourvalueasapersonisbeingjudged.Nomatterhowhardyoutry,youwon'tbeabletoavoidthisone.Noonedoes.Itstemsfromtheopen-endednatureofthethesisproblem.Youhavetodecidewhata"good"thesisis.Athesiscanalwaysbemadebetter,whichgetsyouintoaninfiniteregressofpossibleimprovements.Recognizethatyoucannotproducea"perfect"thesis.Therearegoingtobeflawsinit,asthereareineverything.Settledowntomakeitasgoodasyoucanwithinthelimitsoftime,money,energy,encouragementandthoughtatyourdisposal.Youcanalleviatethisproblembyjumpingalltheexplicithurdlesearlyinthegame.Getallofyourcourserequirementsandexaminationsoutofthewayassoonaspossible.Notonlydoyoutherebyclearthedecksforyourthesis,butyoualsoconvinceyourself,bysuccessfullyjumpingeachhurdle,thatyouprobablyaregoodenoughafterall.2.Nothingelicitsdominantbehaviorlikesubservientbehavior.Expectanddemandtobetreatedlikeacolleague.Thepaperrequirementsaretheexplicithurdleyouwillhavetojump,buttheimplicithurdleisattainingthestatusofacolleague.Actlikeoneandyou'llbetreated\nlikeone.3.Graduateschoolisonlyoneofthetoolsthatyouhaveathandforshapingyourowndevelopment.Bepreparedtoquitforawhileifsomethingbettercomesup.Therearethreegoodreasonstodothis.First,arealopportunitycouldarisethatismoreproductiveandchallengingthananythingyoucoulddoingraduateschoolandthatinvolvesalongenoughblockoftimetojustifydroppingout.ExamplesincludefieldworkinAfricaonaprojectnotdirectlyrelatedtoyourPhDwork,acontractforsoftwaredevelopment,anopportunitytoworkasanaideinthenation'scapitalintheformulationofsciencepolicy,oraninternshipatamajornewspaperormagazineasasciencejournalist.Secondly,onlybykeepingthisoptionopencanyoufunctionwithtrueindependenceasagraduatestudent.Ifyouperceivegraduateschoolasyouronlyoption,youwillbepsychologicallylabile,inclinedtogetabitdesperateandinsecure,andyouwillnotbeabletogiveyourbest.Thirdly,ifthingsreallyarenotworkingoutforyou,thenyouareonlyhurtingyourselfanddenyingresourcestoothersbystayingingraduateschool.Therearealotofinterestingthingstodoinlifebesidesbeingascientist,andinsomethejobmarketisalotbetter.Ifscienceisnotturningyouon,perhapsyoushouldtrysomethingelse.However,donotgooffhalf-cocked.Thisisaseriousdecision.Besuretotalktofellowgraduatestudentsandsympatheticfacultybeforemakingupyourmind.AvoidTakingLectures-They'reUsuallyInefficient.Ifyoualreadyhaveagoodbackgroundinyourfield,thenminimizethenumberofadditionalcoursesyoutake.Thisrecommendationmayseemcounterintuitive,butithasasoundbasis.Rightnow,youneedtolearnhowtothinkforyourself.Thisrequiresactiveengagement,notpassivelisteningandregurgitation.Tolearntothink,youneedtwothings:largeblocksoftime,andasmuchone-on-oneinteractionasyoucangetwithsomeonewhothinksmoreclearlythanyoudo.Coursesjustgetintheway,andifyouarewellmotivated,thenreadinganddiscussionismuchmoreefficientandbroadeningthanlectures.Itisoftenagoodideatogettogetherwithafewcolleagues,organizeaseminaronasubjectofinterest,andinviteafewfacultytotakepart.They'llprobablybedelighted.Afterall,itwillbeinterestingforthem,they'llloveyourinitiative-anditwillgivethemcreditforteachingacourseforwhichtheydon'thavetodoanywork.Howcanyoulose?Thesecommentsofcoursedonotapplytocoursesthatteachspecificskills:e.g.,electronmicroscopy,histologicaltechnique,scubadiving.WriteaProposalandGetItCriticized.Aresearchproposalservesmanyfunctions.1.Bysummarizingyouryear'sthinkingandreading,itensuresthatyouhavegottensomethingoutofit.\n2.Itmakesitpossibleforyoutodefendyourindependencebyprovidingaconcretedemonstrationthatyouusedyourtimewell.3.Itliterallymakesitpossibleforotherstohelpyou.Whatyouhaveinmindistoocomplextobecommunicatedverbally-toosubtle,andintoomanyparts.Itmustbeputdowninawell-organized,clearlyandconciselywrittendocumentthatcanbecirculatedtoafewgoodminds.Onlywithaproposalbeforethemcantheygiveyouconstructivecriticism.4.Youneedpracticewriting.Wealldo.5.Havinglocatedyourproblemandsatisfiedyourselfthatitisimportant,youwillhavetoconvinceyourcolleaguesthatyouarenottotallydementedand,infact,deservesupport.Onewaytoorganizeaproposaltoaccomplishthisgoalis:a.Abriefstatementofwhatyoupropose,couchedasaquestionorhypothesis.b.Whyitisimportantscientifically,notwhyitisimportanttoyoupersonally,andhowitfitsintothebroaderschemeofideasinyourfield.c.Aliteraturereviewthatsubstantiates(b).d.Describeyourproblemasaseriesofsubproblemsthatcaneachbeattackedinaseriesofsmallsteps.Deviseexperiments,observationsoranalysesthatwillpermityoutoexcludealternativesateachstage.Linethemupandstartknockingthemdown.Bytransformingthebigproblemintoaseriesofsmallerones,youalwaysknowwhattodonext,youlowertheenergythresholdtobeginwork,youidentifythepartthatwilltakethelongestorcausethemostproblems,andyouhaveavailablealistofthingstodowhensomethingdoesn'tworkout.6.Writedownalistofthemajorproblemsthatcouldariseandruinthewholeproject.Thenwritedownalistofalternativesthatyouwilldoifthingsactuallydogowrong.7.Itisnotabadideatodesigntwoorthreeprojectsandstarttheminparalleltoseewhichonehasthebestpracticalchanceofsucceeding.Therecouldbetwoorthreemodelsystemsthatallseemtohaveequallygoodchancesonpaperofprovidingappropriatetestsforyourideas,butinfactpracticalproblemsmayexcludesomeofthem.Itismuchmoreefficienttodiscoverthisatthestartthantodesignandexecutetwoorthreeprojectsinsuccessionafterthefirstfailforpracticalreasons.8.Pickadateforthepresentationofyourthesisandworkbackwardsinconstructingascheduleofhowyouaregoingtouseyourtime.Youcanexpectastabofterroratthispoint.Don'tworry-itgoesonlikethisforawhile,thenitgraduallygetsworse.9.Spendtwotothreeweekswritingtheproposalafteryou'vefinishedyourreading,thengiveittoasmanygoodcriticsasyoucanfind.Hopethattheircommentsaretough,andrespondasconstructivelyasyoucan.10.Getatit.Youalreadyhavetheintroductiontoyourthesiswritten,andyouhaveonlybeenhere12to18months.ManageYourAdvisors.Keepyouradvisorsawareofwhatyouaredoing,butdonotbotherthem.Beaninteresting\npresence,notapest.Atleastonceayear,submitawrittenprogressreport1-2pageslongonyourowninitiative.Theywillappreciateitandbeimpressed.Anticipateandworktoavoidpersonalityproblems.Ifyoudonotgetalongwithyourprofessors,changeadvisorsearlyon.Beverycarefulaboutchoosingyouradvisorsinthefirstplace.Mostimportantistheirinterestinyourinterests.TypesofTheses.Neverelaborateabaroqueexcrescenceontopofexistingbutshakyideas.Gorighttothefoundationsandtesttheimplicitbutunexaminedassumptionsofanimportantbodyofwork,orlaythefoundationsforanewresearchthrust.Thereare,ofcourse,othertypesoftheses:1.Theclassicalthesisinvolvestheformulationofadeductivemodelthatmakesnovelandsurprisingpredictionswhichyouthentestobjectivelyandconfirmunderconditionsunfavorabletothehypothesis.Rarelydoneandhighlyprized.2.Acritiqueofthefoundationsofanimportantbodyofresearch.Again,rareandvaluableandasurewinnerifproperlyexecuted.3.Thepurelytheoreticalthesis.Thistakescourage,especiallyinadepartmentloadedwithbedrockempiricists,butcanbepulledoffifyouaregenuinelygoodatmathandlogic.4.Gatherdatathatsomeoneelsecansynthesize.Thisistheworstkindofthesis,butinapinchitwillgetyouthrough.Tocertainkindsofpeoplelotsofdata,eveniftheydon'ttestahypothesis,willalwaysbeimpressive.Atleasttheresultsshowthatyouworkedhard,afactwithwhichyoucanblackmailyourcommitteeintogivingyouthedoctorate.Therearereallyasmanykindsofthesesastheiraregraduatestudents.Thefourtypeslistedserveaslimitingcasesofthegood,thebad,andtheugly.Doctoralworkisachanceforyoutotryyourhandatanumberofdifferentresearchstylesandtodiscoverwhichsuitesyoubest:theory,fieldwork,orlabwork.Ideally,youwillbalanceallthreeandbecometherarepersonwhocantranslatethetheoryfortheempiricistsandtherealworldforthetheoreticians.StartPublishingEarly.Don'tkidyourself.Youmayhavegottenintothisgameoutofyourloveforplantsandanimals,yourcuriosityaboutnature,andyourdrivetoknowthetruth,butyouwon'tbeabletogetajobandstayinitunlessyoupublish.Youneedtopublishsubstantialarticlesininternationallyrecognized,refereedjournals.Withoutthem,youcanforgetacareerinscience.Thissoundsbrutal,buttherearegoodreasonsforit,anditcanbeajoyfulchallengeandfulfillment.Scienceissharedknowledge.Untiltheresultsareeffectivelycommunicated,theyineffectdonotexist.Publishingispartofthejob,anduntilitisdone,theworkisnotcomplete.Youmustmastertheskillofwritingclear,concise,well-organizedscientificpapers.Herearesometipsaboutgettingintothepublishinggame.1.Co-authorapaperwithsomeonewhohasmoreexperience.Approachaprofessorwhoisworkingonaninterestingprojectandofferyourservicesinreturnforajuniorauthorship.He'llappreciatethehelpandwillgiveyoulotsofgoodcommentsonthepaperbecausehisnamewillbeonit.\n2.Donotexpectyourfirstpapertobeworld-shattering.Alotofeminentpeoplebeganwithaminorpieceofwork.Theamountofinformationreportedintheaveragescientificpapermaybelessthanyouthink.Workuptothemajorjournalsbypublishingoneortwoshort-butcompetent-papersinlesswell-recognizedjournals.Youwillquicklydiscoverthatnomatterwhatthereputationofthejournal,alleditorialboardsdefendthequalityoftheirproductwithjealouspride-andtheyshould!3.Ifitisgoodenough,publishyourresearchproposalasacriticalreviewpaper.Ifitispublishable,you'veprobablychosentherightfieldtoworkin.4.Donotwriteyourthesisasamonograph.Writeitasaseriesofpublishablemanuscripts,andsubmitthemearlyenoughsothatatleastoneortwochaptersofyourthesiscanbepresentedasreprintsofpublishedarticles.5.BuyanduseacopyofStrunkandWhite'sElementsofStyle.Readitbeforeyousitdowntowriteyourfirstpaper,thenreaditagainatleastonceayearforthenextthreeorfouryears.Day'sbook,HowtoWriteaScientificPaper,isalsoexcellent.6.Getyourworkreviewedbeforeyousubmitittothejournalbysomeonewhohasthetimetocriticizeyourwritingaswellasyourideasandorganization.Don'tLookDownonaMaster'sThesis.Theonlyreasonnottodoamaster'sistofulfillthegenerallyfalseconceitthatyou'retoogoodforthatsortofthing.Themaster'shasanumberofadvantages.1.Itgivesyouanaturalwayofchangingschoolsifyouwantto.Youcanusethistobroadenyourbackground.Moreover,yourideasonwhatconstitutesanimportantproblemwillprobablybechangingrapidlyatthisstageofyourdevelopment.Yourknowledgeofwhoisdoingwhat,andwhere,willbeexpandingrapidly.Ifyoudecidetochangeuniversities,thisisthebestwaytodoit.Youleavebehindpeoplesatisfiedwithyourperformanceandinapositiontoprovidewell-informedlettersofrecommendation.YouarrivewithmostofyourPhDrequirementssatisfied.2.Yougetmuch-neededexperienceinresearchandwritinginacontextlessthreateningthandoctoralresearch.Youbreakyourselfingradually.Inresearch,youlearnthesizeofasolubleproblem.Peoplewhohavedonemaster'sworkusuallyhaveamucheasiertimewiththePhD.3.Yougetapublication.4.What'syourhurry?Ifyouenterthejobmarkettooquickly,youwon´tbewellprepared.Bettertogoabitmoreslowly,buildupasubstantialbackground,andpresentyourselfabitlaterasapersonwithmoreandbroaderexperience.PublishRegularly,ButNotTooMuch.Thepressuretopublishhascorrodedthequalityofjournalsandthequalityofintellectuallife.Itisfarbettertohavepublishedafewpapersofhighqualitythatarewidelyreadthanitistohavepublishedalongstringofminorarticlesthatarequicklyforgotten.Youdohavetoberealistic.Youwillneedpublicationstogetapost-doc,andyouwillneedmoretogetafacultypositionandthentenure.However,totheextentthatyoucangatheryourworktogetherin\nsubstantialpackagesofrealquality,youwillbedoingbothyourselfandyourfieldafavor.Mostpeoplepublishonlyafewpapersthatmakeanydifference.Mostpapersarecitedlittleornotatall.About10%ofthearticlespublishedreceive90%ofthecitations.Apaperthatisnotcitedistimeandeffortwasted.Goforquality,notforquantity.Thiswilltakecourageandstubbornness,butyouwon'tregretit.Ifyouarepublishingoneortwocarefullyconsidered,substantialpapersingood,refereedjournalseachyear,you'redoingverywell-andyou'vetakentimetodothejobright.AcknowledgementsThankstoFrankPitelkaforprovidinganopportunity,toRayHueyforbeingaco-conspiratorandsoundingboardandforprovidinganumberofthecommentspresentedhere,tothevariousunknowngraduatestudentswhokepttheseideasincirculation,andtoPeteMorinforsuggestingthatIwritethemupforpublication.