On the need for a global academic internet platform

On the need for a global academic in ternet platform Nadja Kutz ∗ Marc h 8, 2008 A CM-Class: K.4.1; K.4.3 Keyw ord s: science, comm un ication, economics, global, computer-supp orted c ollab orative w ork, transb order data flo w, ethics Abstract The article collects argumen ts for the necessit y of a global aca- demic in tern et platform, whic h is organize d as a kind o f “glo bal scien- tific parliamen t”. With suc h a constitution educational an d r esearc h institutions will hav e direct means for comm unicating scientific re- sults, as we ll as a platform for repr esen ting academia and scient ific life in the public. 1 Academic comm unication, rep resen t ation and p o litical p r o ce s ses 1.1 Academic comm unication and its so cietal repre- sen tation In t he last y ears in ternet comm unication has tak en a leading role in ov erall so cietal life. This holds not only true for the w estern w o r ld, but is more and more eviden t on a global lev el as well. New forms of so cial net w orking and so cial comm unities grew within no time, partially furthered b y net w o r king to ols, suc h as wik i’s, blog s, cvs rep os- itories, commercial netw orking sites (e.g. m yspace, f aceb o ok, xing) or other forms o f comm unit y forming platforms reac hing f rom online gaming plat- forms lik e w orld of warcraft, o ver en vironmen ts suc h as second lif e, and online learning platforms to customer services o f online stores. ∗ email: nadja@da ytar.d e 1 P olitical life has partially merged in to this pro cess. Ev ery ma jo r p olit- ical party has a t least a webs it e. P olitical leaders ha v e their o wn w ebsite. In ternational o rganisations ha ve their w ebsites etc. P olitical messages are distributed not only via tra dit io nal media, lik e newspapers, TV stations but more and more often via p olitically colored blogs or directly on media suc h as y o ut ub e. P olitical comm unication platforms such as the W orld Economic F orum [W eF o], F o r a .tv etc. provide meeting and info rmation spaces. Ho we v er academic life, whic h had o nline netw orking tools long b efore the in ternet and whose net w orking to ols (lik e the h tml format, serv er arc hitecture etc) prov ided the gr ounds for the curren t b o om ta k es an astonishingly hidden role in this dev elopmen t. Univ ersities of course hav e their o wn w ebsites. Moreo v er a great deal of academic life ta k es place online. Online registrations, augmen ted learning, studen t netw orks, researc h o v erviews, public a tion lists, lecture notes etc are almost standard at every bigger univ ersity . Moreo ve r univ ersit y mem b ers tak e part in inv estigatio ns or pro vide information for foundat io ns and p olit- ical and ecomomic institutions (lik e the IPCC) and th us they pla y a strong role in the p olitical comm unication pro cess. Ho we v er all these con tributio ns are rather hidden. Ev en in cases where the participation of academic mem- b ers is emphasized these are us ua lly men tioned in diffuse terms lik e “leading climate scien tists” or “experts in genetic engineering”. Another impor t a n t hidden role of academia is the con tribution to kno wl- edge accum ulation within the in ternet. This is not only provide d via the univ ersit y p ortals, but b y the participatio n o f univ ersity mem b ers in collab- orativ e env ironmen ts such as wikip edia. In stark con tr a st to this there is an often strong neglectance o f academia and educational institutions in p olitics. This neglectance tak es on v arious forms. It may b e as direct as budget cuts for r esearc h and educational insti- tutions or it ma y be more sub tle with metho ds reac hing from res training the autonom y of unive r sities, in terference of p o litics in academic pro cesses with ideas lik e “elite formation” t o concrete structural desicions like emplo yment and funding regulatio ns. These p olitical me asuremen ts usually tak e place on a national basis, al- though researc h is hig hly international. The international organizatio ns whic h are dev oted to represen t educa- tional institutions lik e the UNESCO provide informat ion on educational topics and in part also on researc h con t en t. They prov ide to o ls for col- lab orations. How ev er, they are mediators and their mediating role is usually limited, whic h results e. g. in predefined prio rities. Similar things hold true for Science organizations, i. e. they represe n t scien tific life to a certain degree and mediate b et ween academia and so ciet y . 2 This role is imp ortant ho w ev er not exhaustiv e enoug h. In particular the “w eakness” of science organisatio ns to represen t educa- tional institutions has a structural reason. On one hand it is the relativ ely small organisato r ia l size (lik e the UNESCO Sciences Sector has ab o ut 200 staff mem bers (whic h is small if y ou are lo oking for a direct adressee to s et up on a science related question) on the other ha nd it is the v ery role as a mediator whic h diminishes the influence of a science organizatio n. The ab o ve should serv e as a fast explanation that there is and wh y there is a ce rtain lack of a direct activ e represe ntation of academia and academic questions in so cietal life. Suc h a represen tationa l gap could – at least in part – be filled b y an official academic electronic platform, whic h is directly run b y all (or almost all) higher–educational institutions, i. e. univ ersities in the w orld. This exclude s many g o o d think ers and artists but considering only uni- v ersit y mem b ers mak es the authen tification and organization easier. Last not least t he system of univ ersities spans a global net with a rather (emphasis on: rather) high neutralit y to w ards cultural a nd gender sensibilities, a huge exp ertise and a go o d access to lo cal a dministratio ns. 1.2 Academic comm un ication and p olitical pro cesses Besides possibly filling a represe ntational gap a “scienc e parliamen t” migh t emphasize its ro le a s a global consultan t. The purp ose of this section is firstly to briefly recall the structural sens itivit y of p olitical systems and secondly to briefly recall the role of consultan ts in p olitical systems. A profound p olitical analysis is definitely b ey ond the scop e of this article. The reminder should merely serv e as a motiv ation wh y a science parliamen t could act as a consultan t. Demo cratic systems can b e v ery sen sitiv e to rather subtle organisato- rial differences such as b et we en represen tat ive demo cracy , direct demo cracy , b et wee n v oting systems, concerning con trol of p ow er (legislativ e, executiv e, jurisdiction) a.s.o. As an example one can compare the W eimar republic and the curren t german demo crat ic system. It is more or less undisputed that the instability of t he W eimar republic w ere part ially due to its demo cratic organisatorial structures. Of course this has to b e seen in con text with the historical circumstances, but as a matter of fact the W eimar republic sa w 20 cabinet changes in 14 years, whereas the current and former wes t ern Bun- desrepublik of G erman y had 21 c hanges of gov ernmen t in ab out 58 y ear s. Another in teresting p o in t when lo oking at democracies and their repre- sen tational c hara cter is the so cial and psyc hological origin of p o liticians. F o r most democratic systems the so cial mixture of p olit icians do es not mirror 3 the so cial mixture of the corresp onding so ciety . A famous example for the case of Germany is the high p ercen tage of la wy ers in the german go v ernmen t (whic h is apparen tly mostly due to the so called jurist privilege) [DaBu]. Psyc hological pro cesses whic h are in volv ed with raising to and sta ying in p olitical p ow er within a demo cratic sys t em are quite complicated. A p o liti- cian has t o be stress resistan t, stable or a t least emanat ing s tabilit y , resis tan t to in trig ues, b e able to mak e f a st and far reac hing decisions, whic h can ha ve v a st implications. A p olit ician ha s to b e res p onsible etc. In short: a p oliti- cian is a certain kind of a h uman. This implies that an a v erage of polit icians w ould v ery probably act quite differen t ly than an a v erage of the ov erall p op- ulation, which has its adv antages and disadv an t a ges. Gr oup dynamics and for example the confirmation bias whereb y w e seek and find confirmatory evidence in supp ort of a lready existing b eliefs a nd ignore or reinte rpret dis - confirmatory evidence [MSh06 ] are adding another psyc hological comp onen t. Another w ellkno wn fact is that the concrete paths p oliticians c ho ose in their political daily life are oft en to a g reat extend informed b y consultan ts and lobb yists, whic h are mostly represen ting economic forces. It mak es sense to ha v e consultan ts – a p olitician just does not ha v e the time to dig through all the de tails, whic h are of t en needed for a p olitical decision. Ho w ev er as outlined ab ov e t he c hoice of consultants seems t o b e a rather obscure a nd often quite psyc hological issue. It is usually not v ery represen tativ e. This is a n o bvious violation of the idea of a democracy – giv en that there is a demo cracy in a country . According to the so called third T ransformation Index o f the Bertelsmann foundation: Despite the c ontinuing worl d w ide e c onomic gr owth of the p ast few ye ars, mass p overty r emains the c entr al pr oblem in most developing c ountries, and the majority of p e ople have no lasting s h ar e in this pr osp erity. A n d although the numb er o f governments determine d by fr e e ele ctions is gr owing, m any p e ople ar e s til l exclude d fr om p olitic al de cision-ma king or ar e actively denie d other p olitic al and civil rights. This is the so b er- ing c onclusion r e ache d by the thir d T r ansform ation In d ex (BTI), a n international c omp ar ative study of 125 d e veloping and tr ansition c oun- tries.. . [BTI08] Moreo ve r according t o Mr. Janning, globalization ex p ert of the Bertels- mann F oundat io n: ”F r om the glob al p ersp e c tive , advan cing glo b aliza tion is pr o ducing gr e ater over al l gr owth and pr osp erity, but not in a fair or sustainable way. The 4 p ositive effe cts of glo b aliz a tion ar e not b enefiting the majority of p e ople and it is n ot s ustainable for the futur e. The fa ilur e, but also the s o lution to these pr oblems lies in the r eform c ap ability at state and governm ent level.” Summarizing – ne xt to its role of being a represen tativ e of the g lobal sci- en tific comm unity , as had b een outlined in the preceding section, the sc ience platform could a ct as a more o r less neutrally , tra nsparently , and globally acting consultant. It may sa ve guard p oliticians if they ha ve to mak e con- tro vers ial desicions, whic h are facing or will face manho o d so on. It could emp o we r the UN to enforce desicions against lo cal w arlords. Th us it could c hange the p olitical landscape without making apriori sensitiv e structural c hanges at the polit ical systems themselv es, whic h do es not exclude that one k eeps thinking ab out them. Mor eov er it is in principle p ossible to use the “science pa r liamen t” for infor ming structural changes suc h as adaptive managemen t etc. Nev ertheless it should b e emphasized a gain that the main task o f suc h a platform is to serv e as an instrumen t whic h w o uld allow for a b etter represen- tation and co ordination of global academic kno wledge and not as a “shado w go vernme nt”. 2 Scien t ific Met ho ds and the V alidation of Scien t ific Questions 2.1 Scien tific met ho d, kno wledge accum ulation In this se ctio n I would lik e to briefly dis cuss the role of the s cientific metho d and the v alidation of scien tific questions m ainly at the example of math, computer science a nd ph ysics. The purp ose of this is to explain to a nonscien tific audience w h y the desi- cion pro cess in scienc e is differen t from that in p olitics and society . How ever the pro cedure of ho w “t he” scien tific metho d w orks giv es us also indications of how the prop osed in ternet platfor m ma y w ork. Due to the logical nature o f math (the la ng uage for phys ics) the ev aluation of a give n scien tific que stion or h yp othesis is relativ ely straigh tforw a r d. In particular mathematics provide s eve n sometimes notions on w ether a question is solv able at all, on how complex a question may b e or on how random an a nswe r is. Mathematical assertions can b e che cke d for lo gic al c on s istency . Interest- ingly the computer has b ecome more and more imp o rtan t in this in t he last 5 y ears. Assertions in ph ysics can to a great extend b e che cke d by me as ur ements and observations . Ph ysical mo dels/h yp othesis/theories (i.e. the mathemati- cal description of ph ysical en tities) ha v e to b e v alidated in accordance with these measuremen ts/observ atio ns and in accordance with the mat hematics describing them. The whole pro cess mus t b e o b j ectiv e so that the scien tist do es not bias the in terpretation of the results, whic h include s that the measuremen ts/observ a- tions must b e in principle repro ducable in order to verify them. In short there is a quite established widely accepted metho d for c heck ing h yp othesis‘ in mathematics and ph ysics, whic h a pplies t o a great extend also to other sciences lik e biology and c hemistry and par t ially a lso to humanities , lik e so cial sciences and also to economics. The w ay how ev er ho w to set up a hypothesis and the question what questions should b e ask ed is usually not straig h tforward, it is a pro cess whic h in v olv es imagination, in tuition and sometimes also scien tific fashions. 2.2 Scien tific met ho d, review of results The revie w/verific ation pro cess of a result includes v arious fo rmal step s in- cluding the prepublication of res ults, whic h reac hes f rom internal discussions with trusted exp erts to putting them on an electronic arc hiv e sys t em as so- called e-prin ts. Lately there ha d b een some examples, where – mo stly w ell establishe d facult y mem b ers – put draf ts of their scien tific results out for discussion on w ebsites, as a kind of preprepublication. Ho w ev er this presupposes that the w ork had r eac hed a certain stage of maturit y and that the authors are prepared for discussions. The arc hive arXiv.or g , whic h w as founded in 1991 tak es a prominen t role in that, i. e. here almo st all math and ph ysics publications are freely prepublished and sorted in a conten t-classification system. The final step of a publication is then usually done in a p eer review ed journal, where p eer review means that the work is indep enden tly revie wed by usually at least 2 anon ymous e xp erts (the author is usually not anon ymous). The anon ymit y guaran tees to a certain degree that the w o rk is inv estigated solely in terms of conten t and not in terms of things like p ersonal sympa- th y . Whereas it should b e remarke d that is is quite unus ual that “negative results”, i. e. cases where researc h lead e.g. to no result are published at all, although t he description of these cases could constitute v aluable infor ma t io n. This is (v ery) r o ughly what p eople mean by the scien tific metho d o f kno wledge acquisition (please see also the wikip edia p o rtal on “scien tific 6 metho d”). In particular this metho d has been designed to ens ure ob jec- tivit y and designed to a void in particular biases, lik e the ab ov e men tioned confirmation bias. But as men tio ned b efore “the” scien tific me tho d is not a fixed recip e. It is an ongoing cycle, constan tly dev eloping more useful, acc urate and com- prehensiv e mo dels, h yp othesis and metho ds. By the ab ov e it is also clear that this system can ha ve failures, in partic- ular the sensitivit y of the system to funding and rew ards is a delicate issue. 2.3 Scien tific met ho ds , failures, scien tific in tegrit y The purp ose o f this section is to decribe the sensitivit y of the scien tific metho d with resp ect to funding in order to provide an insight in p ossible vulnerabilities of the platform. The sensitivit y of the scien tific metho d to f unding starts with t he c hoice of questions. If researc h has to lead to certain results in a predefined w ay (like via timelines in a researc h prop osal) then questions will b e made with resp ect to wethe r this can be ac hieve d at all or not, whic h implies that questions whic h are presumably to o hard to solve will be left out in suc h prop osals. This holds also true for freq uen t ev aluations, where usually only “p ositive” and “final” ac hiev emen ts are a warded (whic h are in terms of science funding often coun ted as num b er of publications a nd num b er of paten ts), i. e. again – in terms of ev alua t io n y ou b etter choose a sub ject whic h has some chanc e of b eing succes sful. Henc e the shorter and more limited the prop osal/ev alua t ion cycles are, the more results will b e “small” results. Small results or “almost fully satisfying” results can sometimes b e useful e. g. in industrial mathematics, where an in telligen t mathematical optimiza- tion can do sometimes w o nders and ma y alr eady b e a sufficien t progress considering the in ves ted time and money . But think o f how long it to o k to pro ve the F ermat conjecture (ab out 400 y ears) and imagine how man y p eople w ould try to a pply for a gran t prop osal in a similar case. F unding problems c an also w ork as a test case for s cien tific integrit y , i. e. w ether the principles of the scien tific metho d are violated. Th is needs no further ex planation – also scien tists ma y b e corrupt. Ho we ver the scien tific metho ds mak es c orruption m uch harder then in ordinary life. So funding problems result usually rather in unple asan t in teractions among scien tists than in wrong assertions. Ho w ev er funding p olicies ma y distort the o v erall picture, lik e if y ou lo ok fo r evidenc e only in a certain direction then this may lead to insufficien t and ev en wrong conclusions. Despite the usually high integrit y of scien tists the problem of scien tific in tegrity has to b e mentioned – especially in con text of industry/po litically 7 funded exp ertises. T he mor e one individuum or a small r esearc h group is dep enden t o n certain funds the higher is the danger of violation of scien tific in tegrity . Lik ewise this indicates that the mor e diverse and the higher the n umber of in volv ed groups in the discussion of a scien t ific question is, the more in tegrity can b e expected. F urthermore the more op en the pro cess of dev elopping a solution is the more p eer r eview will auto matically ensure mo r e in tegrity . Again here the dep endence on funding/rew a rds ma y result in fears that collegues snatc h aw a y in termediate results etc. and thus in hiding the w ork. F or the case o f the electronic pla t f orm this implies firstly t ha t the running exp enses of the platfo r m has to b e made b y the univ ersities a lone, wether they get reimburs ed b y an ov erall higher budget is another question. An initial ex t r a fundraise to install the tec hnics, etc. how ev er ma y b e useful. Moreo ve r not m uc h direct res earch, whic h dep ends on fundings should b e in v olv ed with the platform (to o exp ensiv e), but available inform a tion should b e r ather gather e d for an ev aluation pro cess. Secondly , discussions should in v olv e as man y work groups as sensible. The more wide-spread and more div erse the groups the more it will also b e hard for lobby ists to influence. Scien tific discussion s hould be as op en as p ossible. Ho w ever it ma y be nec- essary to hide w ork and con tributors fo r prev en ting lobb yism or for other reasons. Exp eriences with infor ma t io n blo c k ades af t er n uclear acciden ts a r e an example, where e. g. po litics in terfered with the pure demand for scien- tific information. T hirdly the sort of questions to b e adressed has to b e o f public in terest, where public may include the scien tific public only . P articu- lar b enefits of companies hav e to b e av oided or at least discussed op enly , as they probably can t be a v oided sometimes, but this holds true in g eneral f o r scien tific results. 2.4 F urther Implications for an electronic platform There is ano t her asp ect one should m en tion. The scien tific metho d deals with scien tific questions. Often the scien tific questions to b e discussed are in stro ng relation to e. g . economical, juridicial and ethical ques tio ns. A natural-scien tific judgemen t whic h inv olv ed the scien tific metho d ma y need to b e ev aluated o r juxtap o sed in terms o f considerations with resp ect t o (eco- nomic, political) realizabilit y and ethics. An example: The use o f genetically mo dified plan ts ma y impose sev ere health risks. W e may come the p oint where one ha s to use genetically mo dified plan ts in order t o feed the planet. (It is not necessary up to no w I think!) So this question has to b e discussed in conjunction with these constrain ts or at least juxtap o sed to them. The h umanities sections of univ ersities are a v ery v aluable partner in doing this. 8 This is wh y the platform could also further interdiscip linarit y . 2.5 A w orkfl o w prop osal for the e lectronic platfo rm The general sc hemes of ho w the scien tific metho d w orks giv es us indications ab out the p ossible design of the pro p osed in ternet platfo rm/net w ork. The exact tec hnical realization o f suc h a platf orm is indeed a sensitiv e issue and b ey ond the scop e of this article. Here a propo sal fo r a general w orkflow sc heme: 1. Define questions The notion of the platform w orking as a “parliament” means that there are predefined questions. Thes e questions come from societ y and science it self. Just lik e la ws or so cietal questions are discus sed in a parliamen t. So so ciet y is basically doing the “what-question-should-b e-to-b e-a nsw ered-searc h” pa rt of the sc ientific metho d. The “ parliamen t mem bers” or “exp erts” are faculty mem b ers of univ ersities. The parliamen t itself is run by univ ersities. It would actually be go o d if the compilation of questions w ould b e preprocessed and discusse d by a f orum whic h is op en to ev eryb o dy , j ust lik e sa y wikip edia. The questions whic h are suitable f or in ves tigation need to b e stated in a precise manner, i. e. scien t ists ma y hav e to reform ulate them or dissect them in terms of scien tific v a lidation, and economical, political and moral sub questions. The questions relev ance has to b e established, it has t o b e ensured that particularism is av oided and it ha s to b e decided w ether a question will b e made in to an official question and a s suc h published on the platform. 2. Supply exp ertise and data Exp erts need to supply data to a g iv en official question, whic h means a v ail- able scien tific material. This presupp oses an initial choice o f experts, whic h ma y sup ervise the gathering of material and of further exp erts. Hence this pro cess is similar to the work of an editor of a journal, who assigns comm u- nicating facult y and these in the turn assign r eferees f o r a w ork. If g lobal exp erts are electronically registered and when their expertise is classified via k eyw ords, lik e in the Mathematics Sub ject Classification [MSC] then the e x- p ert retriev al is fairly simple. Ho w ev er the v alidation of a question may not necessarily b e confine d to exp erts. Non exp erts could a prio ri ha ve the p ossi- bilit y to con tribute, at least by commen ting and pro viding data. Often e. g. studen ts are v ery well if not b etter informed and may con tribute at least 9 for data gathering. It sh ould b e p ossible to inv ite exp erts from outside the univ ersit y facult y esp ecially e. g. members of researc h institutes. If ques tions ar e related to iss ues of sp ecific nations then it has to b e decided w ether this is a nationa l or ev en more lo cal question. If this is not the case it has to b e decided w ether t here has to b e a nationality balance in accordance to may b e UN prop ortions. 3. F orm ulate answ ers/hypothesis‘ Based on their data/exp ertise exp erts may form ulate answ ers (hypothesis) and “prepub lish” them in a “library” nex t to a discussion forum corresp ond- ing the question or if this is necessary assert that no hypothesis can be made, as e. g. there is not enough material etc. Dep ending on the question, con- ferences may b e needed (lik e in the case o f the climate c hange discussions). Here NGOs ma y pla y an imp o rtan t role. 4. Ev alua te answ ers, publish them Based on the discussion preliminary or final answ ers can b e for mulated a nd officially published as suc h. The answ ers can b e explained via the gathered data. The presen tation of the results will certainly need a go o d collab ora- tion with science comm unicators/j o urnalists in order to av oid comm unication problems lik e for example it happ ened in the rep ort ab out c hildho o d cancer in t he vicinit y of n uclear p ow er pla nts [Ra Ch08] In the generic case together with the answ er it should b e made visible ho w many exp erts are in f av our for whic h answ er a nd to what extend. This is what I w ould call a “v ote” or “p oll”. It ma y b e tha t experts decide tha t some asp ects a re more imp ortant than ot hers. The exact questions of how to p oll and decide on the results has to b e decided b y the exp erts, how ev er there could b e simplified v oting mechanis ms. In particular simple questions could b e decided t o b e answ ered via v ot ing on m ultiple c hoice questions. This may sound crude especially in context of the careful design of the scien tific me tho d, but it may often b e sufficie n t for certain questions or at least for intermediate dec isions, lik e ab out the issue of relev ance etc. F or a nice in tro duction to v o t ing see e. g. the AMS math a warene ss mon th w ebsite [AM S08] Dep ending on the question, the v o t e as w ell as the electronic discussion of results itslf m ust in principle bear the p ossibilit y to b e made anon ymous in o rder to sa veguard the in volv ed scien tists as p ointe d out earlier. 10 5. Set timeline A t imeline for further in ves t ig ations and v alidations has to b e giv en next to the “answ ers”. It should be discussed w ether and ho w one could adjust thes e to strategies lik e adaptiv e managemen t. So in principle the w orkflow of the platform is not so muc h differe n t from the scie ntific da y-to- da y practise in that its workflo w resem bles the w orkflo w of the scien tific metho d. Ho we ver there are diffe rences to the day-to-da y practise. I w ould like to emphasize some of them, as w ell as emphasize some other relev ant p o ints: • Th e pla t form would collect and sort structural da t a from all univ er- sities w orldwide and thus provide a w orldwide academic net w ork in a electronic-seman tically conne cted w a y . • Bes ides b eing a n organisatorial framew ork the platform will hav e a task, namely to represen t the global net work of univ ersities and to pro vide answ ers to scien tifically difficult so cietal questions. The questions will come – at least in part – from so ciet y , i. e. in particular the “ a nsw ering service” b y suc h a platform will b e seen as a service of academia to so ciet y . • Th e amoun t of answ ers and the w ork whic h will b e in v olv ed with them is freely a djustable. I. e. if univ ersities dont hav e enough resources they may decide to terminate the service. Lik ewise if so cieties are not con ten t with the service they will pro ceed in cutting do wn science budgets. • Th e answ ers to the give n questions will in principle a lready b e existing, and not researc hed i. e. the answ ers should reflect the curren t scien tific kno wledge rather than constitute res earc h. I. e. the main v alue of the platform is that experts pro vide and connect information and expertise, rather than that they do researc h. This do esn’t exclude of course that this in v o lves small short term rese a rc h o r that further researc h ma y b e necessary (see timeline). • Th e plattfor m could b e used as a call-in instrumen t. If scien tists ar e concerned ab out certain questions, they could call in collegues rather easiliy . Sin ce ev erything is electronic, these calls c an b e simp ly catego- rized according to relev ance, lo cal connectivit y etc. Th us mailing lists could b e assem bled v ery easily . • Sinc e the infrastructure of univers ities is used (computers, ro oms), the cost can b e k ept relative ly small. 11 • Mos t information whic h is needed for the platform is already existing, electronically a v ailable information (like lists of facult y me mbers, e- prin ts, open access journals etc.). This info rmation needs to a g reat exten t only b e connected. This is more a tec hnical ch allenge than an organisatorial o ne. In general, o rganisatorial regulatio ns should b e held minimal and scien tists should b e trusted in their abilit y of self-con tro l and self-organisation (giv en acceptable living and w orking conditions). The expertise of platforms like that of the UNESCO esp ecially w it h their unit win netw orks [UNITWIN] or organisations lik e sense-ab o ut- science [SAS] and other org anisations a re very v aluable a nd one should think ab out how to include them into the pro cess. 3 F urther p oss ible tasks of t h e platform The b elo w section is mostly inte nded to encourage discussion, ab out the organisatorial structures of economies and their relatio n to the mec hanism of assigning v alues, which is usually called “ pricing”. In particular the topics in this section are intended for thinking ab o ut w ether the scien tific platform could serv e as a to o l to inform ab out actual costs and/or indications for pricing. This section con tains a rather simplistic approach to a complicated sub- ject, how ev er simplifying often helps to shed ligh t on the main constituen ts of a system. On the other hand simplifying to m uc h may rather blur the in v olv ed ma in mec hanisms. In o rder to explain what is meant by that and what could b e the moti- v a tion for setting up suc h a “pricing table”, it will b e necessary to make a little excursion on what is inv olv ed when a v alue – a price – is assigned to a go o d. In par ticular it will b e necessary to discuss main for ms of v alues in order to get an understanding of the main mec hanisms of pricing and th us serv e as an explanation on wh y the curren t pricing mec hanisms in finance are insufficien t (at least in m y point of vie w) and ho w they could b e enhanced b y a platform, as prop osed ab ov e. 3.1 Assigning v alues In order t o study the pro cess of ev aluation or pricing o ne should dis cuss the notion o f a v alue. F or brevit y b oth, the name of a v alue and its actual quan tity will b e called a v alue. It should b e clear fr o m the con text what is mean t. 12 In the follo wing I will distinguish b etw een three differen t t yp es of v alues. I will call them measura ble v alues , whic h are in o pp osition to rather emo- tional va l ues suc h as what will b e called sub jectiv e v alues and rational sub jectiv e v alues . measurable v alues If y ou ha v e a siev e and a b o wl of sand then the large sand gra ins sta y in the siev e and the small ones drop through, i. e. one can order the grains accord- ing to their size and hence the here in v o lv ed v alue is called size. Moreov er there was no need to in v en t this v alue as it w as ph ysically already imma- nen t. So a measuremen t (in the example the measuremen t of the grain size) is a pro cess where a v alue is assigned to something, i. e. it is a kind of ev alua t ion. In the case of the measuremen t this ev aluation is do ne via a ph ysical pro cess (sieving). In the fo llo wing I would like to call also processes, whic h can be mathematically quan tified, but whic h dep end on a set of v al- ues/measureme n ts and the correspo nding v alues as me asur e d v a lues r elative to something . So e. g. if I ha ve tw o grains of differen t w eigh ts then I call their comp ound w eigh t a measured v alue relativ e to the give n individual weigh ts. (Lik ewise on could see the grain size as a measured v alue in relation to the size of the siev e holes ) . How ev er I do not w an t to fix this not io n to o m uch . The impo rtan t p oin t here is that for the dete rmination of t he v alue w e mak e use of scien tific procedure which is more or less exact rep eata ble and describ- able. Nev ertheless it should b e p oin ted out that the c hoice of what and ho w is also related to a tten tion. I. e. if we stic k to the example of the sandgrain then the differen t sizes of the sand particle caugh t our at t ention and made us prepare the experiment with our sand and siev e in this particular w a y . This is no bad thing p er se but it is imp ortant when talking ab out ob jectivity in scien tific reasoning. Go o d scie n tific reasoning tries to reduce the att en tion factor. sub jectiv e v alues and rational sub jectiv e v alues Another metho d of a ssigning v alues is b y a sub jectiv e a nd mostly emotional judgemen t, lik e e. g. to assert that one lik es one p erson more than the ot her or b oth equal etc. means to assign a subje ctive v a l ue . Here it is ev en harder to sep era t e the c hoice of assigning a v alue from the issue of atten tion especially if these emotions are link ed to ev o lut io nary needs. Ho w eve r it is p o ssible to a certain degree. In particular v alues whic h were agreed up o n within a comm unit y/collectiv e are in some sen se sub jectiv e, ho wev er they are moder- ated b y the collectiv e know ledge. F or t ha t reason I would like to call thes e 13 kind of mo derated sub jectiv e v alues r ational subje ctive values . Among oth- ers they a re intended for detac hing t he sub jective emotional ev aluation of an individual from highly fluctuating mo o d c hanges. An example: If someone stomps on your feet, y our reflex could b e to stomp back , ho we v er usually y ou would decide t o ma yb e turn aw ay g rudgingly b ecause you don t w an t to b eha ve badly . A constitution consists to a great part of o r is based on rational sub jec- tiv e v alues. As sertions lik e w ether death p enalt y is acceptable or not a re rational sub jectiv e v alues. There is no scien tific metho d to measure w ether killing someone should ha v e the v alue go o d or bad. Some comm unities state that suc h v alues were assigned b y deities (lik e e. g . “you shall not kill” in the bible) ho w ev er they are still something whic h had been agreed on b y a comm unit y , so one can ke ep calling t hem rational sub jectiv e v alues. Th us an imp ortant feature of r ational subje ctive v alues is that they ar e often w ritten down somew h er e . The ab ov e displa ys in particular that a s o cietal “ rule” ma y be seen as a “v alue”, in that the corr espo nding so ciety /collectiv e had agreed up on that this rule w as “go o d” or “accepte d as a law”. Aga in here the name o f a v alue (namely that what is measured whic h is: ho w go o d do you fulfill a rule) and its actual quan tity (quantifie d v alue: go o d) will b e for brevit y both b e called a v alue. side remark: Conversely a me asur ement c an b e se en as a rule, in the c ase of the sieve the rule would b e imp ose d by the physic al r e ality, i. e. t he rule c ould b e “you have to fi t thr ough the sievehol e s ” (or the opp osite rule not to fit thr ough). Or in other wor ds: by observing wether a sand p article fits thr ough the sieveholes me ans to observe wether it f ulfil ls the rule “you have to fit thr ough the sieve- holes” or n ot. As a matter of fact: if one ac c epts this view of a m e asuring app ar atus b eing a “rule” this c ould imply that the c onstancy of a v a lue may b e r elate d to the question of “how go o d the rule” has b e en “written d o wn ”. Or in other wor d s : h ow many similar app ar atus’ ar e ther e/c an b e built in or der to m e asur e a p articular value in question. As an il lustr ation: I f ther e is onl y one p articular ch unk of a gauge kilo gr am (which is curr ently in Paris) a n d this kilo gr am is actual ly changing its size (wh i c h it actual ly do es), then the actual value of a p articular de facto c o nstant mass is changing ac c or dingly (for that r e ason physicists ar e now lo ok i n g for another standar d to me asur e mass). But if ther e would have b e en sever al identic al k i l o gr ams fr om the b e- ginning on then on aver age this “ga uge kilo gr am de c ay pr o c ess” MA Y have b e en sl o wer, i. e . the c onstancy b etter. Inter estingly this wo uld give a kind 14 of “unc ertainty r elation ” b etwe en the r ate of cha nge of “on av e r age me asur e- ment ac cur acy” and “sp ac e” (i. e. the numb er of on aver age ide al c opies of me asur ement app ar atus). The motiv at io n o f a collectiv e t o establish rational sub jectiv e v alues is – as already indicated – in order to mo derate b etw een t he sometimes ra t her immediate v alues of the individuum a nd v alues whic h conce rn the collec- tiv e, lik e e. g. how to make a collectiv e surviv e or to work more efficie ntly , whic h often includes nonemotional measurable v alues suc h as size of water ressources etc. Ho w efficien tly common goals of a collectiv e are ac hiev ed is related to t he no t io n of “trust”, i. e. to t he question of how rules/rational sub j ectiv e v alues a r e accepted in the corresp onding collectiv e [RaT ru08]. So the term r ationa l subje ctive value is al w ays in r elation to a c ol le ctive . rational v alues, the collectiv e and common goals Ho we v er a collectiv e is usually part of a bigger collectiv e. If the size of the smaller collectiv e is small in comparision to the bigger collectiv e (lik e a family in a nation) then the v alues of the smaller collectiv e shall for simplicit y also b e regarded as individual sub jectiv e with resp ect to the bigger collectiv e. So families or companies ma y b e regarded as individua in relation t o a na t io n, i. e. they ar e likely to adopt the nations v alues. If the groups are rather big g er (lik e e. g. migrants) than their v alues usually hav e to b e regarded sep erately , but t his can also be done with respect to a nations v alues, i. e. they ma y add differen t v alues or reev aluate a nations rational sub jectiv e v alues. Rational sub j ectiv e v alues can thus b e seen as a kind of a verage v alue, av eraged in terms of p opulation and time, but still b elonging to the individuum (like in terms of a nation these c ould be called cultural v alues, in terms of a religious group, religious v alues, in terms of consumers, consumer v alues etc.) Under generic c onditions one can assume t ha t a collectiv e b eha ve s mostly according to the giv en ph ysical circumstances (measurable) and their collec- tiv e desicions (based on the rational s ub j ective v a lues, la ws, rules). How ev er as p o in ted out ab o ve this b ehaviour is blurred, disturb ed by the sub jectiv e individual v alues, dep ending on the size and nature of so cietal/collective con trol. A collectiv e c ho oses their v alues often in order to ac hiev e a set o f common collectiv e g oals. E. g. the rational sub jectiv e v alue of s o me c hristians to explain c hristianit y , to ev angelize is go o d ma y lead to the common collec tiv e c hristian goal to make ev eryb o dy turn in to a chris tian. An imp ortant p o in t here is that it is somewhat p o ssible to determine to what extend a c hoice of ratio nal sub jectiv e v alues meets a common c ollectiv e goal or in other words: a collectiv e rational sub jectiv e v alue. This can b e 15 done b y observ ation. Another example: if a society holds the rational sub- jectiv e v alue: killing is not go o d (y ou shall not k ill) and controls this rational sub j ectiv e v alue e. g. via la w enfor cemen t then one can o bserv e w ether this reduces the o v erall killing rate (a collectiv e rational sub jectiv e v alue) in that so ciet y . Often collectiv e rational sub jectiv e v alues comp ete with eac h other so it is usually nev er p o ssible to re ac h one common goal. (like if a religious gro up decides to adv ertize their religion in a nonforceful obtrusiv e manner, as their common goal is not t o force people in t o their b eliev e then this could mean that they ma y gain less follow ers). So in short: rational sub jectiv e v alues are somewhat in tended to me et collectiv e rational sub jective v alues in an optimized w a y . (computer) mo dels based on v alues In terestingly t wo imp ortan t phy sical quantities enter here again, namely time and space. Space enters via the q uestions of ho w the individ ua are connected and ha ving their v alues compete with eac h other (the connectivit y and the amoun t of individua has a g eog raphical/spatial comp onent to it), whereas time en ters in that it is needed for observing the collective pro cess, whic h includes the formation of rational s ub jectiv e v alues and lik ewise the observ a- tion w ether ra tional sub jectiv e v alues are in accordance with e. g. a collectiv e rational sub jectiv e v a lue. An imp ortant fact is that the durat io n and size of the inv olv ed time and space can b e abstractly reduced b y mathematical computations w hic h model the in v o lv ed system including its v alues and the pro cedural connections (de- p endencies) b et w een the se v alues. But of course the mathematical model can o nly approximate the real outcome. F or the accuracy of a mathematical prediction the correct assessmen t of all individual v alues as w ell as all measurable v alues and their in terplay is essen tial. F or large collections of v alues this usually turns out to b e feasible if one restricts the sample size and iden tifies the main mo des of interaction. Lets call t his a c hoice of sample. So the accuracy of the mathematical prediction relies to a great part o n this c hoice of sample and the stability of that choice ov er time. Roughly one can sa y that this mathematic al choic e of sample is e asier and mor e likely to b e ac cur ate if the involve d values an d their interplay c an b e me asur e d or when they are fixed in la ws/rules etc. i. e. if they ar e r ational subje ctive values . This is wh y complicated phys ical predictions (like abo ut climate c hange) 16 can b e quite accurate (a s t hey include mostly measurable v alues), whereas predictions whic h in volv e sub jectiv e v alues, like this is often the case for economic v alues are usually no t so accurate. I. e. the more an ev alua tion is less rational (in the ab ov e sense) like in panic buy outs at sto c k markets or lik e in the recen tly observ ed desaster a t the societe generale, the harder it is to determine the main inv olv ed v alues and their interde p endencies. In econom y the ev aluation (t he assignmen t of v a lues) is usually done via pricing (I will call incen tive s such as interes t rates etc. and o t her economic judgemen ts also pricing in order to make t hing s simple r ) . So for ex a mple the price of a resource dep ends on the a moun t of t he phys ically accessible part of that ressource (measureable ), the mining regulations in a country (rational sub j ectiv e v a lues) and e.g. the p o lit ical situation (these are sub jectiv e v alues as they imply that no collectiv e agreemen t on v alues had b een ac hiev ed). Lets lo ok at ano t her example, whic h – due to recen t ev ents (the mov ing of a Nokia plant from German y t o Romania) – may b e w orthwhile to in ves tig ate, namely the v alue of la b our. In a collectiv e, lik e a nation the v alue of lab our dep ends among o thers on • sk ill (time and qualit y of education needed fo r that skill, a moun t of p eople whic h are in principle capable of acquiring that skill) [more or less measurable v alue] • edu cational resources (amoun t of capable people who can b e trained) [measurable v alue] • basic living costs (relative t o giv en prices , lik e to a bask et of av ailable commo dities (W a renk orb)) [measurable v alues in relatio n to the bask et] • geographical flexibilit y (this applies t o employ er as w ell as to emplo y ee: a fa r mer can hardly mov e, if his soil is under drought, whereas an agent in a Call Cen ter is not b ound to lo cation, ho wev er this flexibility may dep end also on rational sub jectiv e v alues suc h as langua ge (journa ls, journalists, la wy ers etc.) a nd immigration/principal deplo ymen t la ws. The geographical flexibility a lso determines ho w m uc h lab our costs are in dep endencies to lab our costs in other coun tries. [depending on ra- tional sub j ective a nd sub jectiv e v alues] • negotiation (the prices are under negotiation b et w een the emplo ye r and the emplo y ee. These dep end on cultural (lik e gender and ethnicit y biases) and p olitical p ow ers (trade unions, exclusivit y of skill etc.)) [mostly dep ending o n rational sub jectiv e v alues] 17 • political (nat ional and in ternat io nal p olitics may subsidiz e certain sec- tors, ov erall conditions ma y b e unstable (whic h secretely enhances costs) etc. [dep enden t on rational sub jective and sub jectiv e v alues] The ab o v e example illustrates again ho w mostly sub jectiv e v alues mak e the pricing into a complicated issue. Consequen tly failures in economic predictions are most lik ely less due to o ve r work ed so called “quants” (mathematicians/phy sicists in finance) but more lik ely due to the fact that economy inv olv es non-rational judgemen ts/ v a lues. (As a side remark: It is illustrativ e to discus s the NOKIA example based on these criteria, since it ma y displa y t hat romania n w o rk ers are not nec- essarily w orking h arder than germans for the money , but that mostly the living conditions are partially che ap er and t ha t the so cial costs/inv estmen ts are lo w er. In part icular it is probable t hat the w orking conditions (w o rk hours, v acation t ime) are not as go o d as in German y (part of negotiation), ho we v er this point most lik ely do esn’t accoun t for the v ast difference s in w age. Moreov er it dis pla ys that p eople with the skills required for w o r king in that NOKIA facility are relativ ely easy to find. In addition due to high mo ving costs, immigration, language, cultural and so cial conditions (and so- cial problems) and to the probably a little smaller W arenk orb (less g o o ds can b e purc hased f or the w age) presumably not to o man y germans will mov e to Romania in order to keep working for NOKIA.) Usually the ab o ve kinds o f analysis’ are done b y e. g. traders/ec onomical analysts. Ec onomical decisions are made up on their judgemen t and ability to ass ign price s. This is also wh y they are usually rather w ell pa id. This fact b y the wa y ho lds true since the b eginning of trade. A go o d trader was able to (more or less) correctly determine prices (usually a ssigned to a geographical lo cation) in order to kno w where to buy and where to se ll. Of course part of the job of a sales/trade p erson is also negotiation, but t his c ha ng es a price usually only partially – the main p oint is the abilit y to correctly dete rmine and analyse the price s (last not least the negotiatio n itself dep ends on this abilit y). v alue types and t he parad o x of v alue It is illustrativ e to discuss the no t io n of utility (i. e. the inc rease of p o sitive v a lues and the decrease of negativ e v alues) in terms of the ab o v e in tr o duced t yp es of v alues. Lik e for example the so called diamond-w a t er parado x (also called parado x o f v alue), which basically states that it is absurd that t he price of diamond is so muc h higher then the price of w ater, b ecomes clearer if one 18 ac know ledges the follo wing: Let us assume that there is enough abundance of water for eac h individuum in a collectiv e (sinc e scarcity could rise the prise of w ater in to arbitrary he igh ts, as it is a necessit y fo r liv ing b eings) and that there is no other use for diamond than b eing used as a jew elry item. Then the utilit y of diamond could b e v ery differen t f or the v arious in- dividua in that collectiv e. In particular some p eople may ev en dislik e a diamond’s color a nd shap e etc. Hence the sub jectiv e v alues/prices for dia- monds could apriori greatly differ dep ending on the individuum. Ho we ver if a greater share of that collectiv e agrees that sa y the nice refra ctio n prop er- ties of diamond are something that has a p ositiv e v alue (and thus a higher utilit y) then this implies tha t the rational sub jectiv e v a lue/price of diamond will b e higher. This rational sub jectiv e v alue may b e related to individual sen tiences lik e that the refraction giv es a comforting feeling etc. but it need not to b e related apriori. This is also where branding b ecomes imp ortan t in that it influences the formation of rational sub jectiv e v alues. Scarcity ma y amplify a collectiv e agreemen t on price. Lik e if it is easy to pro duce dia- mond in great amoun ts then the price w ould probably appro ximately b e the pro duction/distribution costs, but as it is scarce the price is higher. On the other hand if a collectiv e agrees that diamond has no or small v a lue, then probably only a few p eople would dig for it. This is more or less also the r eason why the europ eans who en tered America tra ded gems against glass p erls: the rational sub jectiv e v alues o f Americas nativ es regarding the v a lue of glass vers us gem where different then the ones of Europ ean natives , whic h was of course partially due to the f a ct that Americas nativ es didnt kno w ab o ut glass pr o duction. It should also b e men tioned that b esides b eing a jewe lry it em diamond usually includes of course other uses. In particular due to its scarcit y , solidit y and recognizabilit y/recall v alue it may serv e as medium of exc hange. v alue assignmen t and tradabilit y The correct assignmen t of prices can get arbit r a rily complicated and so math- ematicians use for that problem a wkw ard sounding to ols lik e risk calculus of v on Ne umann - Morgenste rn or Arrow - Pratt risk a v ersion. The ab ov e should th us only serv e as an intro ductory ex p osition ho w science en ters so cietal a nd economical questions. There exists v arious platforms for financial analysis’ lik e e. g. Blo o m b erg L.P . whic h pro vide bac kground info rmation (including p olitical) and soft ware to ols. Lik ewise professional analysts and e. g. rating agencies ar e adding their exp ertise to the financial and economical w orld. So wh y should a scien tific pla tform provide a similar service? 19 As w as already p ointed out b efore a lot of the pricing mec hanism de p ends on the abilit y of tra ders to analyse and judge ab o ut prices. Here a funn y feedbac k happ ens. In particular the accuracy of pricing dep ends also on the prop ert y of b eing tradeable. Or in o ther w ords if there is no tra ding reason for assigning a price to something then this missing price will destort the actual price. This explains wh y suc h sub jectiv e v alues, whic h are often ev en rational sub jectiv e v alues for natio ns (!) lik e a health y en vironment for future generations or human so cial conditions are not necessarily rational sub jectiv e v a lues for tr a ders. This implies that t he pitfalls of capitalism are also due to that org anisatorial problem of an inaccurate and/o r mis sing pricing of v alues b ey ond the trading scop e. It mak es the thinking ab o ut capitalism less in to an ideological question. Ho we v er it is o f course p ossible to assign v a lues/prices b ey ond the tr a ding scop e. This has e. g. b een demonstrated in the so called Stern r eview, where the future costs of climate c hange w ere estimated. How ev er these prices are only sparsely reflected in no wada ys econom y and trade. Similar estimations could be ma de for n uclear en ergy . Nuclear energy is curren tly (esp ecially with resp ect t o regenerativ e energies) relativ ely c heap, ho we v er j udg ed b y en vironmen tal cos ts a nd future en vironmen tal costs (lik e o ve r the next 2 0 000 y ears) it is very exp ensiv e [this is a measurable v alue]. So the actual traded price for n uclear energy is simply one thing: wrong. Lik ewise the almost “zero v alue” of digitized items , lik e e. g. the o ne of digital music are in fact almost zero v alue only in terms of trading. It is not true that these items ha v e no v alue or no costs. Eve ry m usician w ould ag ree on this. Th us it is not to o far fetche d to conjecture that these items are not so in teresting for the trading mark et, b ecause the value exch a nge for digital items , i. e. t he tr ad i n g of digital items is – due to e. g. priv ate cop ying – almost non-manageable. This holds also true for e.g. digital storage s pace. Here the actual costs a re usually acclaimed somewhe r e else, as was already p oin ted out by Chris Anderson [And08] whic h he - among o thers - illustrated at the example o f Go ogle. This mec hanism is a bit comparable to dumping prices, just that dumping prices need not to b e due to a non-manag ibilit y o f the tra ding pro cess. As a matter of fa ct non-tradability tak es also pla ce in the case o f a monop ole, where the actual price and the real costs are often in a misre- lation. Concluding: There is a need to analyse the price of g o o ds also with r esp ect to apriori non tradable v alues, lik e p ollution, so cial conditions, future living conditions etc. As already explained an economic platfo r m will do this only to a limited extend . Institutes like the New Economics F oundation [NEF] 20 ha ve usually only rather limited resources to pro vide such a servic e. Small institutions ma y also b e less neutral. A global plat f orm, w here economy / financial mathematics departmen t s of univ ersities could bring in their exp ertise, where agriculture experts, climate scien tists, cultural exp erts etc. work together would ha v e a differen t qualit y . As a side effect suc h a service could ev en hav e an economic v alue, as e. g. insurance companies often need to find s ubstitutes f o r these pric ing informations in things lik e future risks. This ma y encourage on the other hand the financial w orld t o donate financial informa t io n to the prop osed in ternet platfor m. Last not least a more detailled and true analysis of prices w ould allo w for p olitical me asuremen ts similar to the emission tradings. One could e. g. in tro duce a similar thing f o r the f ut ure costs of nuclear w aste, for defo r esta- tion etc. One could introduce certificates for maxim um work hours, health b enefits to emplo y ees etc. (a lthough legal regulations a r e lik ely to b e more efficien t in some of these cases). A more corr ect and scien tifically sound determination of a price is a necessary condition for that. As a lr eady indicated ab o ve the actual “how” of the determination of a v a lue is very complicated and b ey ond the scop e of this article. Often it is already helpful to understand whic h main pricing pro cesses are inv olv ed and ho w they dep end o n eac h other. The notion ”determination of a price” cer- tainly includes “fuzzy” determinations. Or in other w or ds it is already helpful if a big round of exp erts could a ssess economical/so cietal etc. price ranges and approximate future dev elopments a nd th us a ssist in making complicated p olitical decisions, lik e e. g. in the case of mandated marke ts. A scien tifically minded determination of a price could ev en tually a lso include game-lik e struc tures lik e betting on prices, a kind of toy sto c kmarke t etc. Also if I do not agree with Robin Hanson that “Betting mark ets are our b est know n institution for aggr ega ting info r ma t ion.” [Ha] b eha v ourial strategies like barg a ining, bluffing, risk av ersion etc. (see e.g. [Ca03]) play an imp o r t a n t role in economics and last not least in the determination of a price. 4 Ac kno w ledgements I w o uld like to thank Sabine Hossenfelder for v aluable help. Our comm uni- cation [Bee08] can b e found in t he comment section to a p ost on the blo g randform on whic h I prepublishe d most of the though ts, whic h w en t in to this article. It is also on her blog Back reaction, where I found the link t o senseaboutscience [SAS]. 21 I found the link to Robin Hansons article [Ha] on Scott Aa ronsons blog.[Aa] I would like to thank Tim Hoffmann for a critical r eading of the do cumen t. References [Aa] http://scot taaronson.com/blog/?p=309 [AMS08] http://www. amstat.org/mathandvoting/index.cfm? fuseaction= Main [And08] http://www. wired.com/techbiz/it/magazine/16- 03/ff_ free?curren tPage=1 [Bee08] http://www. randform.org/blog/?p=1639#comments [BTI08] http://www. bertelsmann- trans formation- i ndex.de/ [Ca03] Colin F. Camerer, Beha vioral game theory , Exp erimen ts in Strategic In teraction Prin c eton University Pr ess 2003 [DaBu] Datenhandbuc h zur Gesc hic h te des Deutsc hen Bundestages : http://www. bundestag.de/bic/dbuch/index.html see a lso table in wikip edia: http://de.w ikipedia.org/wiki/Bundesregierung_ %28Deutschl and%29#Anteil_der_Volljuristen_in_der_ Bundesregie rung [Ha] http://hans on.gmu.edu/futarchy.html [MSC] http://www. ams.org/msc/ [MSh06] Mic hael Shermer “The p olitical bra in” Scientific Americ an Magazine . July 2006. [NEF] http://www. neweconomics.org/ [RaCh08] http://www. randform.org/blog/?p=1629 [RaT ru08] http://www. randform.org/blog/?p=1614 [SAS] http://www.s enseabouts cience.org.uk/ [UNITWIN] http://port al.unesco.org/education/en/ev.php- URL_ID= 41557&URL_D O=DO_TOPIC&URL_SECTION=201.html [W eF o] http://www. weforum.org/en/index.htm 22