Authors Daniel Mietchen[1], Gregor Hagedorn[2], Konrad U !"rstner[#] [$] , M !abiana Kubke[%], &laudia Kolt'enburg[(] [)] , Mark Hahnel[*] and +yubo,ir -ene.][/] 1 0 1cience #2 2 0 3ulius K4hn56nstitut, K"nigin5+uise51tra7e 1/, 1$1/% 8erlin, Ger,any # 0 6nstitute 9or Molecular 6n9ection 8iology, Uni.ersity o9 W 4 r'burg, D5/)2*2 W 4 r'burg, Ger,any $ 0 :esearch &entre 9or 6n9ectious Diseases, Uni.ersity o9 W 4 r'burg, D5/)2*2 W 4 r'burg, Ger,any % 0 1chool o9 Medical 1ciences, Uni.ersity o9 ;uckland, <e= >ealand ( 0 &ellular ?herapy and ?ransplantation ) 0 &linic 9or 1te, &ell ?ransplantation, Uni.ersity Medical &enter Ha,burg5@ppendor9, Ger,any * 0 1cience #2 / 0 -enso9t -ublishers, 1#a Geo Mile. 1treet, 1111 1o9ia, 8ulgaria Abstract 1cienti9ic research is a process concerned =ith the creation, collecti.e accu,ulation, conteAtuali'ation, updating and ,aintenance o9 kno=ledge Wikis pro.ide an en.iron,ent that allo=s to collecti.ely accu,ulate, conteAtuali'e, update and ,aintain kno=ledge in a coherent and transparent 9ashion Here, =e eAa,ine the potential o9 =ikis as plat9or,s 9or scholarly publishing 6n the hope to sti,ulate 9urther discussion, the article itsel9 =as dra9ted on 1pecies56D 5 httpBCCspecies5idnetD a =iki that hosts a prototype 9or =iki5based scholarly publishing 5 =here it can be updated, eApanded or other=ise i,pro.ed Introduction "Science is based on building on, reusing and openly criticising the published body of scientific knowledge." --Panton Principles [1] ;t its inception, a research proEect is typically 9ueled by at least one o9B ne= data, ne= ,ethodological approaches, or ne= hypotheses @ach o9 these can contribute either to s,all incre,ental steps that 9urther speci9ic kno=ledge or ha.e ,ore pro9ound e99ects on the theoretical 9ra,e=ork o9 a 9ield ?he generation or interpretation o9 data is there9ore constantly changing Despite the tools and ,ethods ,ade possible by the =eb, the scienti9ic literature continues to consist ,ainly o9 Eournal articles, book chapters, ,onographs, preprints and con9erence proceedings, arranged in a se,i5hierarchical structure that .aries considerably across 9ields While such persistent, reliably unchanging kno=ledge eApressions are an essential ingredient o9 the process o9 science, the dyna,ic nature o9 science in general calls 9or dyna,ic publishing 9or,ats that can be updated in light o9 ne= research results ; side e99ect o9 static publishing is the 9rag,entation o9 the scienti9ic landscape, in itsel9 a ,aEor access barrier to accessing scienti9ic content on the =eb Fther i,portant access barriers result 9ro, pay5=alls erected around the docu,ents containing research reports or the =orks cited therein ?he considerable ,o,entu, reached by the Fpen ;ccess ,o.e,ent 5 in both its Green and Gold shades 5 is likely to alle.iate this proble, signi9icantly o.er the neAt 9e= years Ho=e.er, e.en i9 the =hole scienti9ic N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1
:
P o s t e d
6
A p r
2 0 1 1 literature =ere to beco,e 9reely accessible, a nu,ber o9 traditions =ould continue to stand in the =ay o9 e99icient scholarly co,,unication ?he current publishing standard centered around the scienti9ic GpaperG H=ith introduction, ,aterials and ,ethods, results and discussion sectionsI leads to a high le.el o9 redundancy bet=een papers on related topics ; large portion o9 ,ost publications Htypically the introduction and parts o9 the discussionI ser.e to Jrecapitulate eAisting kno=ledgeJ ?hese recapitulations tend to be .ery 9rag,entary =ithin indi.idual articles, yet highly repetiti.e =hen aggregated across articles !or instance, a Google 1cholar search 9or the phrase JMagnetic :esonance 6,aging is aJ[2] returned about 2%22 results at the ti,e o9 =riting, and ,ultiple other phrases are in use to eAplain the .ery sa,e concept to si,ilar audiences 6t =ould it be less tiring 9or authors and re.ie=ers 5 and ,ore con.enient to readers 5 to ha.e 9e=er o9 these introductions, =hile at the sa,e ti,e ha.e better collaborati.e introductions that are under ongoing re.ision :e.ie= articles already pro.ide these 9eatures, and i9 ,ade a.ailable .ia Fpen ;ccess, they could be e99iciently eAploited to a.oid redundancies and to 9acilitate critical appraisal and replication instead ?he traditional scienti9ic paper is designed to put together a series or collection o9 9indings in order to tell a ,ore co,plete story ?his structure inter9eres =ith the ability to e99iciently re9er to Hand thus, later, to retrie.eI each o9 those indi.idual 9indings !or eAa,ple, citations typically go to a =hole publication, an entire ,onograph, a =hole chapter or page, instead o9 to a ,ore precise point o9 interest =ithin the larger publication, that is actually the source 9or a particular clai, With the increase o9 ,ulti5author publications, each contributing to s,all portions o9 a co,plete report, such a type o9 re9erencing results in inadeKuate attribution to the creatorCs o9 the kno=ledge 6n addition, the current citation syste, in =hich re9erences are listed as endnotes or 9ootnotes instead o9 being directly linked 9ro, the citing passages, beco,es a burden to the reader and discourages the reading o9 the original literature Microcitation speci9ic to indi.idual 9igures, data sets or paragraphs are generally discouraged by printed ,edia on the grounds that they increase printing costs ?he increased e99iciency o9 linking to s,all grained in9or,ation bits, ho=e.er, has been pro.en on the =eb 6n the standard publishing en.iron,ent, updating any in9or,ation reKuires the generation o9 a ne= article Gi.en the gro=ing nu,ber o9 research groups and the rapid production o9 ne= in9or,ation, static articles can Kuickly beco,e outdated :e.ie= articles ,ay already be outdated by the ti,e they 9inally get published 6n the case o9 print copies, updating in9or,ation is rarely econo,ically .iable 9or acade,ic publications other than teAt books !or teAt books, ironically ho=e.er, the process o9 updating is especially lengthy, resulting in a long delay and o9ten large portions o9 outdated ,aterials 6n an online publishing conteAt, producing an update can be ,ore easily i,ple,ented and such a step co,es at a lo=er cost ?his ,akes the =eb a ,ediu, that is ,ore aligned =ith the dyna,ic reKuire,ents o9 scienti9ic co,,unication !urther, online publishing o99ers tools 9or .ersion control and the ability to display, co,pare and cite di99erent .ersions o9 a gi.en docu,ent :euse59riendly open content licenses can be used to lo=er sharing barriers enabling kno=ledge producers and users to handle cross5docu,ent .ersioning in an e99icient ,anner Good .ersion control tools support Hand are likely to pro,oteI collaborati.e editing and updating &ollaboration on science reporting ,ight also be pro,oted i9 ne= types o9 author attribution are put in place that link not only large chunks but also s,all5si'e contributions to indi.idual authors <e= 9indings can potentially be added as Gstand aloneG sections =ithin the eAisting dyna,ic entity, or optionally =ith links to other pages =ith ,ore detailed in9or,ation Hsuch as electronic lab notebooksI While radically di99erent 9ro, the =ay research reporting is being per9or,ed in ,ost 9ields today, a collaborati.e and dyna,ic =ay to rapidly update eAisting kno=ledge is better suited to the needs and ti,e scales o9 kno=ledge generation =ithin a typical research proEect ;n updated perspecti.e on scienti9ic output reporting sche,es =ould ,ean that J=ritingJ Hand hence JpublishingJI should actually consist o9 collaborati.e updating "Although the potential of the nternet for i!pro"ing the disse!ination of infor!ation is now N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1
:
P o s t e d
6
A p r
2 0 1 1 taken for granted, it would see! that the attitudes of those in"ol"ed in sharing this infor!ation ha"e not kept pace with the technology." --#he P$oS %edicine &ditors '()1)*[+] Wikis as dynamic publishing environments ?he issues outlined abo.e ha.e pro,pted the suggestion o9 a set o9 J&riteria 9or the in9or,ation scienti9ic Eournal o9 the 9uture[$]J ?herein, =ikis are eAplicitly ,entioned as J=orking plat9or,s o9 the 9utureJ &onseKuently, it appears logical that the criteria are no= collaborati.ely re9ined on a =iki en.iron,ent[%] Wikis beha.e di99erently than ,ore con.entional publishing plat9or,s =ith respect to the 9ollo=ing criteriaB they are .ersioned, their contents is o9ten a.ailable under reuse59riendly licenses, they allo= recapitulation to be replaced by conteAtuali'ation through hyperlinks to background articles, they support 9ine5grained re9erencing and attribution, and are =ell suited 9or transparent re.ie= and ti,ely re.isions Moreo.er, they are J=eb5nati.eJ and thus 9ully co,patible =ith se,antic enhance,ents and the inclusion o9 ,ulti,edia With respect to dyna,ic publishing ,odels, three approaches go in this directionB 1 @Atending traditional publishing by pro.iding the article in an editable 9or,at in addition to the standard non5editable ones, 2 @Atending the traditional repository approach by supplying pre.iously published Fpen ;ccess articles =ith an editable .ersion, and # Mo.ing the entire publishing process to a =iki plat9or, Lirtually all scholarly articles published today are a.ailable in -D! 9or,at 6n addition to that, H?M+ .ersions are a.ailable 9or ,ost, and MM+ .ersions 9or so,e Eournals @Aperi,ents =ith other 9or,ats 5 including =ikis 5 continue He g, .ideos at the 3ournal o9 Lisuali'ed @Aperi,ents[(]I ?he ,ain purpose o9 pro.iding a =iki publishing ,odel is to allo= published articles to beco,e editable, so as to allo= the, to stay current and be eApanded as kno=ledge e.ol.es :endering a published article editable reKuires legal per,issions 9or copying and ,odi9ication Under the pre.ailing copyright arrange,ents, that gi.e publishers a long ter, Htypically around 122 yearsI ,onopoly on the scienti9ic articles, editable =iki .ersions o9 entire articles published by the ,aEority o9 conte,porary scholarly Eournals are prohibited by copyright la= While editable .ersions o9 articles behind pay=alls are certainly o9 li,ited use, ,aking so,e aspects o9 the article a.ailable under a reuse59riendly license ,ay =ell be =orth considering as a disse,ination strategy Gi.en that so,e ,ethodologies and data cannot be copyrighted under ,ost Eurisdictions, those sections o9 the article could be released under ,ore open licenses ; related approach is taken by the Eournal :<; 8iology[)], =hich reKuires[*] article sub,issions regarding ne= :<; 9a,ilies to be acco,panied by the dra9t o9 a corresponding Wikipedia entry Hthat =ould then cite the original paperI ?his ne= policy has been =ell recei.ed by the co,,unity, to the eAtent that a nu,ber o9 authors ha.e since sub,itted dra9ts o9 Wikipedia articles on topics 9or =hich the Eournal does not reKuire that[/] Fpen ;ccess Eournals ha.e a =ider range o9 options to consider =hen rendering their publications editable ?hey could, 9or instance, acti.ely upload the i,ages and other non5teAt ,edia to collaborati.e ,edia repositories like Wiki,edia &o,,ons, place the entire teAt or parts thereo9 5 or a synopsis, i9 a.ailable 5 in a =iki en.iron,ent 69 they do not =ish to engage in these acti.ities by the,sel.es, the open licenses enable others to do part or all o9 this instead Ho=e.er, the handling by the Eournals itsel9 can be ,ore e99icient and auto,ated ?his syste, =orks best i9 the =iki en.iron,ent does not yet ha.e an entry on the topic in Kuestion, as is the case =ith no.el :<; 9a,ilies in the eAa,ple cited abo.e ?he description o9 taAono,ically ne= species is another such situation, and =e ha.e no= i,ple,ented a =ork9lo= that allo=s the descriptions N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1
:
P o s t e d
6
A p r
2 0 1 1 o9 ne= ani,al species published in the Eournal >ooKeys[12] to be uploaded onto 1pecies 6D[11], a =iki dedicated to taAono,y and host to the =iki .ersion o9 the article you are reading !or eAa,ple, the =iki entry on the di.ing beetle <eobidessodes dar=iniensis sp n[12] H&oleoptera, Dytiscidae, 8idessiniI 9ro, northern ;ustralia =as set up the sa,e day that the original description[1#] had appeared in >ooKeys, and 9our =iki contributors ha.e since edited it to i,pro.e and enhance its presentation While this page =as started ,anually, an auto,ated =ork9lo= to i,port taAon treat,ents 9ro, 9uture >ooKeys ,anuscripts into the =iki is currently being 9inali'ed H-ene. et al, >ooKeys, upco,ingI Wiki entries can also be set up retroacti.ely 9or suitably licensed articles already published in other 9or,ats 6n taAono,y, this =ould be especially interesting 9or legacy taAon treat,ents 9ro, do'ens or hundreds o9 years ago that are o9ten barely accessible Fther scenarios are possible, 9or instance, an article on ?hirdhand 1,oke @Aposure &oncerns[1$] in the Eournal @n.iron,ental Health -erspecti.es[1%] has been adapted into a corresponding =iki entry at the @ncyclopedia o9 @arth[1(] ?echnically, there are 9e= li,itations to i,porting scholarly contents into =iki repositories, so it could be en.isaged that =iki .ersions ,ight one day co,ple,ent other 9or,s o9 scholarly co,,unication -rototypes 9or such scenarios already eAist 5 so,e science blogs[1)] run on =iki so9t=are, and a nu,ber o9 electronic lab notebooks are being kept public by =ay o9 =iki plat9or,s, eg at FpenWetWare[1*] or Use9ul &he,istry[1/] With o.er (2 222 &&58N5licensed articles currently a.ailable 9ro, -ubMed &entral in MM+ 9or,at, a syste,atic approach to i,porting such content into a =iki stands eAcellent chances o9 seeding a use9ul tool 9or literature searches, especially i9 co,bined =ith se,antic enhance,ents Hsee belo=I ?he eAa,ples ,entioned abo.e concerned the integration o9 =ikis into publishing =ork9lo=s 6t is also possible to handle the entire Eournal =ork9lo= on a =iki, as eAa,pli9ied by the re.ie= Eournal 1cholarpedia[22] 1uch an approach naturally 9its =ith public peer re.ie= Hie the re.ie=s are being posted in publicI, and the re.ie=ers ha.e Hunless the authors eAclude thatI the possibility to edit the sub,itted teAt by the,sel.es rather than =riting a report about =hat should be changed ?his adds a ne= Kuality to the peer re.ie= process ?he idea o9 a =iki Eournal has also sur9aced 9ro, ti,e to ti,e at Wikipedia and other =iki proEects, but no decisi.e action has been taken yet ;s in the case o9 the Wiki :epository, there is no inherent reason to belie.e that such an approach =ould only =ork 9or 9or,al publications Hor re.ie= articles e.enI 5 this =iki5style peer re.ie= by direct editing could in principle be i,ple,ented 9or any step o9 the research cycle, and be it lab notebooks or grant proposals 1ocial kno=ledge organisation on =ikis ,irrors to a large eAtent the processes o9 scienti9ic publications Fn plat9or,s like Wikipedia, dedicated H.oluntaryI editors process the inco,ing ,aterials, si,ilar to ho= it is done in ,anaged Eournals Material that is undesirable or clearly belo= Kuality standards is re,o.ed o9ten =ithin ,inutes HOuick deletion =ork9lo=I Fther substandard ,aterial is sub,itted to a one5=eek process o9 pointing out errors and reKuesting i,pro.e,ents and re.isions H:eKuests 9or deletion =ork9lo=I ?o so,e eAtent, this process in.ol.es eApert re.ie=, pro.ided eAperts ha.e organi'ed the,sel.es appropriately Heg J:edaktionenJ on Ger,an Wikipedia 5 J@ditorial tea,sJ are under discussion[21] 9or the @nglish .ersion, tooI While the eApert re.ie= in Wikipedia ,ay occasionally be substandard, the ,ass re.ie= is highly transparent, and all i,pro.e,ents resulting 9ro, it i,,ediately enhance the Kuality o9 the product &rucial ele,ents o9 site policies 9or =iki5based publishing are guidelines on =hich kinds o9 content should be uploaded or ,odi9ied and by =ho,, =ho is ad,itted to the plat9or,, =hat are the nor,s o9 interaction, and, 9or instance, to =hat eAtent auto,ated approaches =ill be per,itted !or instance, WikiGenes[22] ,onitors -ubMed 9or ne= gene5related content and posts eAtracts thereo9 on the =iki, =hereas Gene Wiki[2#] har.ests ,olecular biological databases to keep in9oboAes in Wikipedia articles updated N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1
:
P o s t e d
6
A p r
2 0 1 1 Despite their great potential to radically i,pro.e the searchability as =ell as the auto,ated analysis and processing o9 the corpus o9 scienti9ic literature, the uptake o9 se,antic technologies in scholarly publishing is slo=, and 9or the 9e= Eournals that already do pro.ide se,antically tagged content, possible reuse scenarios ha.e yet to be eAplored Manual annotation is labour5intensi.e, and auto,ated annotation not yet reliable outside =ell5circu,scribed areas ;gain, taAono,y is in a pri.ileged position here, and both >ooKeys as =ell as 1pecies 6D can handle tags obeying the ?aA-ub[2$] eAtension to the <ational +ibrary o9 Medicine o9 the U1; H<+MI Docu,ent ?ype De9initions HD?DI[2%] standard Outlook ; syste, o9 the kind described abo.e =ould gra.itate to=ards a s,aller nu,ber o9 ,ore ,eaning9ul and use9ul resources on any gi.en topic ;s readers, scientists =ould be able to 9ocus on a s,aller nu,ber o9 publications, trusting that they contain the ,ost up5to5date in9or,ation, that e99iciently link to the rele.ant data, processes, and re9erences ;s =riters, scientists could choose to contribute to the collaborati.e e99orts reducing the ti,e spent on =riting long repetiti.e ,anscripts that distract the, 9ro, the task o9 generating ne= kno=ledge Fne i,portant incenti.e to =rite traditional publications is the reputation authors gain in the process 6n the current syste,, the .alue o9 a publication is ,ost o9ten associated =ith the percei.ed .alue o9 the Eournal in =hich the =ork is published Hi,pact 9actor, 6!I or through the nu,ber o9 citations an author recei.es HH59actor, H!I 8ut reputation resulting 9ro, both 6! and H! 9all eKually on all authors o9 a gi.en publication Micro5contributions H9or eAa,ple, in the Gene Wiki proEectI, =hen aggregated, can positi.ely bene9it the scienti9ic co,,unity and they should too incorporated into the reputation port9olio o9 scholars F:&6D[2(] 5 a uniKue identi9ier 9or researchers =hich is under de.elop,ent =ith strong support 9ro, ,any scienti9ic publishers 5 could e.entually pro.ide the architecture to support the accu,ulated ,easure,ent o9 a scientistGs acti.ity, taking into account both traditional publications as =ell as ,icro5contribution ,odels Historically, one o9 the ,ost .aluable i,pro.e,ent pro.ided by the traditional Eournal is the role played by pre5publication peer re.ie=ers 8ut the peer5re.ie= syste, does not scale to the realities o9 the scienti9ic 9lo= o9 the 21st century and post5publication peer5re.ie= do not 9eed back into the syste, 1cienti9ic reputation syste,s could ,ake use o9 end5user rating syste,s Has part o9 their post5publication peer re.ie=I that are increasingly 9ound in =eb ,edia He g, 1tack F.er9lo=[2)]I and that can easily reach across disciplines ?he reputation syste,s described abo.e are o9 crucial i,portance, since they do,inate the syste,s through =hich researchers are assessed at the ti,e o9 applying 9or Eobs or pro,otions, tenure, research 9unding and other i,portant ele,ents and ,ilestones o9 the indi.idualGs acade,ic career -er9or,ance assess,ent syste,s could easily acco,,odate to these ne= 9or,ats, assessing data contributions Hthe currency o9 scienceI instead o9 GpapersG Hthe currency o9 scholarly publishersI ?his reKuires e99icient interaction =ith data repositories He g Dryad[2*] or !ig1hare[2/]I Wiki en.iron,ents o99er the opportunity 9or co,,enting on and rating o9 the article content by ,eans o9 associated JtalkJ or JdiscussionJ pages 6n contrast to the classical non5public anony,ous pre5publication peer re.ie= currently in place at ,ost Eournals, the public peer re.ie= process on =iki plat9or,s allo=s re.ie=s to be read by any interested party, and i9 re.ie=ers choose to re.eal their na,e, they can take credit 9or their contributions ; transition to a =iki style5based publication syste, is ideally aligned =ith the realities o9 scienti9ic =ork9lo=s, allo=ing indi.idual contributions to trans9er to scienti9ic reputation and assess,ent syste, in a ,ore suitable granular ,anner Conclusions Wikis can be used 9or dyna,ic scholarly publishing in se.eral =ays, ranging 9ro, scenarios =hich N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1
:
P o s t e d
6
A p r
2 0 1 1 change eAisting publishing =ork9lo=s in ,inor =ays to a 9unda,ental restructuring thereo9 ?he =ider uptake o9 such collaborati.e approaches hinges upon co,,unity acceptance and proper recognition o9 contributions to large5scale collaborati.e en.iron,ents You can edit this document![3! "e#erences 1 0 -anton -rinciples -rinciples 9or Fpen Data in 1cience 2 0 httpBCCscholargoogleco,CscholarPKQJMagneticR:esonanceR6,agingRisRaJ # 0 ?he -+o1 Medicine @ditors H2212I 3ournals, ;cade,ics, and -ande,ics -+o1 Med )H%IB e12222*2 doiB121#)1CEournalp,ed12222*2 $ 0 &riteria 9or the in9or,ation scienti9ic Eournal o9 the 9uture % 0 Wiki.ersityB3ournal o9 the 9uture ( 0 3ournal o9 Lisuali'ed @Aperi,ents ) 0 :<; 8iology ho,epage * 0 :<; 8iology Guidelines 9or ;uthors / 0 -lans 9or :9a, 221252211 12 0 >ooKeys ho,epage 11 0 1pecies 6D ho,epage 12 0 <eobidessodes dar=iniensis entry at 1pecies 6D 1# 0 <eobidessodes dar=iniensis entry at >ooKeys 1$ 0 8urton ; 2211 Does the 1,oke @.er :eally &learP ?hirdhand 1,oke @Aposure :aises <e= &oncerns @n.iron Health -erspect 11/Ba)25a)$ doiB1212*/Cehp11/5a)2 1% 0 @n.iron,ental Health -erspecti.es ho,epage 1( 0 ?hirdhand 1,oke @Aposure &oncerns @ncyclopedic article at the @ncyclopedia o9 @arth 1) 0 -ost by 3i, : Wilson to the MediaWiki5l ,ailing list, 3ul 1, 222* 1* 0 FpenWetWare ho,epage 1/ 0 Use9ul &he,istry ho,epage 22 0 1cholarpedia ho,epage 21 0 -roposalB@ditorial tea,s H:edaktionenI, Wiki,edia 1trategic -lanning =iki 22 0 WikiGenes ho,epage 2# 0 Gene Wiki ho,epage 2$ 0 ?aA-ub page at 1ource9orge 2% 0 <ational +ibrary o9 Medicine o9 the U1; H<+MI Docu,ent ?ype De9initions HD?DI 2( 0 F:&6D ho,epage 2) 0 1tack F.er9lo= ho,epage 2* 0 Dryad ho,epage 2/ 0 !ig1hare ho,epage #2 0 &lick here to edit the docu,ent :eKuires registration at 1pecies 6D N a t u r e
P r e c e d i n g s
:
h d l : 1 0 1 0 1 / n p r e . 2 0 1 1 . 5 8 9 1 . 1