LU D W IG M IS E S

S O C IAL S C IE N C E

AN D

N ATU R AL S C JE N C E

R eprinte iX lromJounN ar, oF S ocIAL P nrlosopnv & JunrspnuosN cr

V ol. 7 , N o. 3 , AP ril 1942

S O C IAL S C IE I{C E AN D

N ATU R AL S C IE N C B

Luowle M rsrs

I

TH E foundations of the modern socia l sciences were la id in theI e ighteenth century. U p to this time we frnd history only. O f course,

the writings of the historians are full of implications which purportto be va lid for a ll human action irrespective of time and milieu, andeven when they do not explicitly set forth such theses they necessarilybase the ir grasp of the facts and the ir interpretation on assumptions

of this type. But no attempt was made to clarify these tacit supposi-

tions by specia l analysis.O n the other hand the be lie f prevailed that in the fie ld of human

action no other criterion could be used than that of good and bad.If a policy did not a tta in its end, its fa iiure was ascribed to the mora linsufficiency of man or to the weakness of the government.

'W ith good

rnen and strong governments everything was considered feasible .Then in the e ighteenth century came a radica l change. The founders

of P olitica l E conomy discovered regularity in the operation of themarket. They discovered that to every state of the market a certa instate of prices corresponded and that a tendency to restore this statemade itse lf manifest whenever anything tried to a lter it. This insightopened a new chapter in science. P eople came to rea lize with astonish-ment that human actions were open to investigation from other pointsof view than that o{ mora l judgment. They were compelled to recognize

a regularity which they compared to that with which they were a lreadyfamiliar in the fie ld of the natura l sciences.

S ince the days of C antillon, H ume, the P hysiocrats and AdamS mith, economic theory has made continuous-a lthough not steady-

S O C IAL S C IE

progress. In the course ofthan a theory of market op

on private ownership oftime been a genera l theor

preference.

The e lements of socia l ,

any concrete images thatmake them easier to visualca l language. For some

popular. There were write :lous extremes. It will suffir

Today the mechanistic n

ca l basis for its applicatitsocia l science. Fositivism b

and economics taught. H .

economics a specia l kind c

postula tes a socia l science,rnethod as idea lly appliedexperimenta l, mathematicabecause science is measui

verification by facts.E very proposition o{ thi

The socia l sciences in ge

based on experience in th

natura l sciences. S ocia l ex

every experience is the ex1

tinguishes socia l experienc

natura l sciences is that it i

phenomena. The experienc,

success is the experiencedifferent e lements of chanl

the conditions of change

of assigning to each effecl

phiiosophica l problem inv

'C f . f^or instance his La P athola

240

S O C IAL S C IE N C E AN D N ATU R AL S C IE N C E 24Iprogress. In the course of this development it has become much morethan a theory of market operations within the frame of a society basedon private ownership of the means of production. It has for sometime been a genera l theory of human action, of human choice andpreference.

n

llre elements of socia l cognition are abstract and not reducible toany concrete images that might be apprehended by the senses. Tomake them easier to visualize one likes to have recourse to metaphori-ca l language. For some time the biologica l metaphors were verypopular. There were writers who overworked this metaphor to ridicu-lous extremes. It lviil sufiice to cite the name of Lilienfe ld.t

Today the mechanistic metaphor is much more in use. The theoreti-ca l basis for its application is to be found in the positivist view ofsocia l science. Fositivism blithe ly waved aside everything rvhich historyand economics taught. H istory, in its eyes, is simply no science;economics a specia l kind of metaphysics. In place of both, Fositivisrnpostulates a socia l science which has to be built up by the experimenta lmethod as idea lly applied in l\ewtonian physics. E conomics has to beexperimenta l, mathematica l and quantita tive . Its task is to measure ,because science is measurement. E very statement must be open toverifrcation by facts.

E very proposition of this positivist epistemology is wrong.The socia l sciences in genera l and economics in particular cannot be

based on experience in the sense in which this term is used by thenatura l sciences. S ocia l experience is historica l experience. of courseevery experience is the experience of something passed. But what dis-tinguishes socia l experience from that which forms the basis of thenatura l sciences is that it is a lways the experience of a complexity ofphenomena. The experience to which the natura l sciences owe a li the irsuccecs is the experience of the experiment. In the experiments theclifferent e lements of change are observed in isolation. The control ofthe conditions of change provides the experimenter with the meansof assigning to each effect its sufficient cause. W ithout regard to thephilosophica l problem involved he proceeds to amass o'facts.,,

These

'C l. Ior instance his La P athologie S ocia le , paris, 1896.

242 JO U R N AL O F S O C IAL P H ILO S O P H Y AN D JU R IS P R U D E N C E

facts are the bricks which the scientist uses in constructing his theories.They constitute the only materia l a t his disposal. H is theory must notbe in contradiction with these facts. They are the ultimate things.

The socia l sciences cannot make use of experiments. The experiencewith which they have to deal is the experience of complex phenomena.They are in the same position as acoustics would be if the only mate-ria l of the scientist were the hearing of a concerto or the noise of awaterfa ll. It is nowadays fashionable to style the statistica l bureauslaboratories. This is misleading. The materia l which statistics pro.vides is historica l, that means the outcome of a complexity of forces.The socia l sciences never enjoy the advantage of observing the conse-quences of a change in one e lement only, other conditions be ing equal.

It follows that the socia l sciences can never use experience to verifythe ir sta tements. E very {act and every experience with which t}rey haveto deal is open to various interpretations. O f course, the experience ofa complexity of phenomena can never prove or disprove a statementin the way in which an experiment proves or disproves. W e do not haveany historica l experience whose import is judged identica lly by a llpeople . There is no doubt that up to now in history only nations whichhave based the ir socia l order on private ownership of the means ofproduction have reached a somewhat high stage of welfare and civi-liza tion. N everthe less, nobody would consider this as an incontestablerefutation of socia list theories. In the fie ld of the natura l $ciencesthere are a lso differences of opinion concerning the interpretation ofcomplex facts. But here freedom of explanation is limited by thenecessity of not contradicting statements satis{actorily verified byexperiments. In the interpretation of socia l facts no such limits exist.E verything could be asserted about them provided that we are not con-fined within the bounds of principles of whose logica l nature we intendto speak la ter. H ere however we a lready.have no mention that everydiscussion concerning the meaning of historica l experience imper-ceptibly passes over into a discussion of these principles without anyfurther reference to experience. P eople may begin by discussing thelesson to be learnt from an import duty or from the R ussian S ovietsystem; they will very quickly be discussing the genera l theory of inter-regional trade or the no less pure theory of socia lism and capita lism.

The impossibility of experimenting means concomitantly the im-possibility of measurement. The physicist has to deal with magnitudes

S O C IAL S C I]

and numerica l re la tions, binvariable re la tions betwrment provides him with tIn human behavior therestandard which could be :

ments which could establirW hat the statistician es

prices and supply or betwpofiance only. If he deterrof potatoes in Atlantis inlowed by a fa ll in rhe priabout what happened or n

potatoes in another countras that of e lasticity of demmeasurement, e .g., specificbody rea lizes that the behaother commodity is variathings in a different way,same individual with chanlviduals in classes which remine the conditions whichcumstances we have to rea .torian and not an experitconstitutes a method of his

In every science the conof an equation are of a noof the equation has a practions which it includes areis possible to introduce sper

mine those unknown. These

logica l designing; they are ranalysis but a lso the startinpwhere there are no constantions are void of practica l a

qualms concerning the ir fothat they are without any pr

But the chief objection Itreatment of economic probl

S O C IAL S C IE N C E AN D N ATU M L S C IE N C E 243

and numerica l re la tions, because he has the right to assume that certa ininvariable re la tions between physica l properties subsist. The experi-ment provides him with the numerica l va lue to be assigned to them.In human behavior there are no such constant re la tions, there is nostandard which could be used as a measure and there are no experi-ments which could estabiish uniformities of this type.

W hat the statistician establishes in studying the re la tions betweenprices and supply or between supply and demand is of historica l im-portance only. If he determines that a rise of 10 per cent in the supplyof potatoes in Atlantis in the years between 1920 and 1930 was fol-lowed by a fa ll in the price of B per cent, he does not say anythingabout what happened or may happen with a change in the supply ofpotatoes in another country or a t another time. S uch measurementsas that of e lasticity of demand cannot be compared with the physicist'smeasurement, e .g., specific density or weight of a toms. O f course every-body rea lizes that the behavior of men concerning potatoes and everyother commodity is variable . D ifferent individuals va lue the sarne

things in a different way, and the va luation changes even with thesame individual with changing conditions. W e cannot categorize indi-viduais in classes which react in the same way, and we cannot deter-mine the conditions which evoke the same reaction. U nder these cir-cumstances we have to rea lize that the statistica l economist is an his-torian and not an experimenter. For the socia l sciences, sta tisticsconstitutes a method of historica l research.

In every science the considerations which result in the formulationof an equation are of a non-mathematica l character. The formulationof the equation has a practica l importance because the constant re la -tions which it includes are experimenta lly established and because itis possible to introduce specific known va lues in the function to deter-mine those unknown. These equations thus lie a t the basis of techno-iogica l designing; they are not only the consummation of the theoretica lanalysis but a lso the starting point of practica l work. But in economics,where there are no constant re la tions between magnitudes, the equa-tions are void of practica l application. E ven if we could dispose of a llqualms concerning the ir formulation we would still have to rea lizethat they are without any practica l use.

But the chief objection which must be ra ised to the mathematica ltreatment of economic problems comes from another ground: it rea lly

244 JO U R N AL O T S O C IAL P H ILO S O P H Y AN D JU R .IS P R .U D E N C E

does not deal with the actual operations of human actions but with af;ctitious concept that the economist builds up for instrumenta l pur-poses. This is the concept of sta tic equiiibrium.

For the sake o{ grasping the consequences of change and thenature of profit irr a market econorny the economist constructs a ficti-tious system in which there is no change. Today is like yesterday andtomorrow will be like today. There is no uncerta inty about the future ,and activity therefore does not involve risk. But for the a llowanceto be made for interest, the sum of the prices of the complementaryfactors of production exactly equals the price of the product, whichmeans that there is no room le ft for profit. But this fictitious concept isnot only unrealizable in actual life ; it cannot even be consistentlycarried to its ultimate conclusions. The individuals in this fictitiousworld would not act, they would not have to make choices, they wouldjust vegetate . It is true that economics, exactly because it cannot makeexperiments, is bound to apply this and other fictitious concepts of asimilar type. But its use should be restricted to the purposes which itis designed to serve . The purpose of the concept o{ sta tic equilibriumis the study of the nature of the re la tions between costs and prices andthereby of profits. O utside of this it is inapplicable , and occupationwith it va in.

N ow a ll that mathematics can do in the fie ld of economic studies isto describe sta tic equilibrium. The equations and the indifferencecurves deal with a fictitious state of things, which never exists any-where . T (rhat they a fford is a matirematica l expression of the definitionof static equilibrium. Because mathematica l economists start from thepre judice that economics has to be treated in mathematica l terms theyconsider the study of static equilibrium as the whole of economics.The pure ly instrumenta l character of this concept has been overshad-owed by this preoccupation.

O f course, mathematics cannot te ll us a .nything about the way bywhich this static equilibrium could be reached. The mathernatica ldetermination of the difference between any actual sta te and the equi-librium state is not a substitute for the rnethod by wirich the logica lor non-mathematica l economists le t us conceive the nature of thosehuman actions which necessarily would bring about equilibrium pro-vided that no further change occurs in the data .

O ccupation with static equilibrium is a misguided evasion of the

S O C IAL S C I

study of the main econo

equilibrium concept shoui

for the solution of one 1

e laboration of sta tic equilThe case is similar wi

price of a commodity as

curve of demand and thewe do not know anythinga posteriori the prices, wtion, but lve do not know tthe past. The represenlat:than a didactic means of :

easily comprehensible .The mathematica l ect-rn

as a measurement of va luwe have to say that pricea le the amount of monelnot equiva lent to the conthe buyer va lues the comr

va lues it lower than the p

this fact and to assumein va luation. W hen e itherequiva lent of the commodwe may say every transar

P hysicists consider the

have no knowledge of whof a fa lling stone. But theof the stone in experimenllaws of fa lling. F rom thebuild up the ir theories profrom the concrete to the n

E conomics deals with l

with commodities, econorpower to experiment withourselves, a knowledge o{

S O C IAL S C IE N C E AN D ]\ATU R AL S C IBN C E 245study o{ the rna in economic probrems. The pragmatic va lue of thisequilibrium concept shourd ,r* rr" underrated, rr"", i, i, an instrurnentfor the solution of one problu* orrly" I,, any case the mathematicare laboration of static *quilibriu* is mere 'y-piay in economics.The case is similar with the use of curves. W e may represent tireprice of a commodity -as

the point of intersection oi-i*o curves, thecurve of demand and the curu" of s'ppiy. But we lr*" t" rea lize t'a twe do not knon' anything about the shape of these "*u"r. sle knowa posteriari the prices, which we assume to be the points of intersec_tion, but we do not know the form of the curve e ither in advance or forthe past' The representation of the curves is therefore nothing morethan a didactic -"-"."_ . of rendering t,e theory graphic and hence moreeasily comprehensible .

The mathematica l economist is prone to consider the price e ither

n ii:T ffT ::J # i:: ;"" il,T ;:.T :;"r;""# iltT {I",*;a.e the amount of,money exchanged for_a cornmodity. The price isnot equiva lent to the comrnodity. A purchase takes place only whenthe buyer varues the commodita higher than the price , and the se 'er- va lues it lower than the price . ivob"oay has the right to abstract fromthis fact and to i

i n r.a tr u a r i c,n wh " i':: il;,:x"":;

t,?jffi , Irffi JH :,n ;,i",I:ujequivarent of the commoditv no transaction takes place. In this sensewe mav say every transaction is for both partie . u ,iru. uin.,,

IIIFhysicists consider the objects of the ir study from without. Theyhave no knowledge^of what i. going- on in the interior, in the o,sour,,,

of a {a lling srone. Bur they huul thJ opportunity to observe the faringof the stone in experiments and the"eby to discover what they ca l theIaws of fa liing. F rom the resurts of s,r.h experiment"i trro*tuage theybuiid up the ir theories proceeding r.o; the speciar ,o ,lr" .or" generar,frorn the concrete to th" *or" ul".r"u"i.E conomics dears with human actions, not as it is sometimes sa id, iwith commodities, economic guantities or prices. w'e do not have tre ipower to experiment with human actions. But we hruu, b"irrg hu*u'ourselves, a knowledge o{ what goes on within acting rnen. W e know

246 JO U R N AI O F S O C IAL FH ILO S O P H Y AN D JU R IS P R U D E N C E

something about the meaning which acting men attach to the ir actions.

W e know why men wish to change the conditions of the ir lives. W e know

something about that uneasiness which is the ultimate incentive of the

changes which they bring about. A perfectly satisfied man or a man

who a lthough unsatisfied did not see any means of improvement wouldnot act a t a ll.

Thus the economist is, as C airnes says, a t the outset of his researches

a lready in possession of the ultimate principles governing the phenorn-

ena which form the subject o{ his study, whereas mankind has no

direct knowledge of uitimate physica l principles. H ere in iies the

radica l difference between the socia l sciences (mora l sciences, G eistes-

utissens,clmften) and the natura l sciences. W hat makes natura l science

possibie is the power to experiment; what makes socia l science possible

is the power to grasp or to comprehend the meaning of human action.

W e have to distinguish two quite different kinds of this comprehen'

sion of the rneaning of action: we conceive and we understand.'W 'e conceive the meaning of an action, that is to say' we take an

action to be such. .W e

see in the action the endeavor to reach a goal by

tle use of means. In conceiving the meaning of an action we consider

it as a purposeful endeavor to reach some goal, but we do not regard

the quality of the ends proposed and of the means applied. W e con'

ce ive activity as such, its logica l (praxeologica i) qualities and cate '

gories. All that we do in this conceiving is by deductive analysis to

bring to light everything which is conta ined in the first principle ofaction and to apply it to different kinds of thinkable conditions. This

study is the object of the theoretica l science of human action (praxe '

ology) and in particular of its hitherto most developed branch, eco'

nomics (economic theory).E conomics therefore is not based on or derived (abstracted) from

experience. It is a deductive system, starting from the insight into

the principles of human reason and conduct. As a matter of fact a llour experience in the fie ld of human action is based on and condi-

tioned by the circumstance that we have this insight in our rnind. W ith-out this a pri,ori knowledge and the theorems derived frorn it we could

not a t a ll rea lize what is going on in human activity. O ur experience

of human action and socia l life is predicated on praxeologica l and

economic theory.It is important to be aware of the fact that this procedure and method

S O C IAL S C IE N I

are not peculiar only to scJ

ordinary da ily apprehensior

ciples and the deductions i

professional economist butfacts or problems. The laymdifferent from that of the sc

less scrupulous in examininland therefore sometimes mo

any discussion on currenl ec

turns very soon towards a cr

any reference to experienc,

S oviet system without fa liincapita iisrn and socia iism. Y

without fa lling back on th,

prices, that means the gener

f3s1"-161 us set aside the

such a thing-is open to diffrequire e lucidation by theor

E conomics is not only not

sible to verify its theorems l

of a complex phenornenon,

in different ways. The san

cla imed as' confirrnations of

It is instructive to compal

in the socia l sciences with tlbooks on economics which,

chapters in which an atteml

by an appeal to the facts. Th

takes. H e sta l'ts from facts

his theory in using them. Ifa state of a ffa irs not yet diikind of experiment would b

to be verified if the result co

radica ily and significantly,socia i sciences.

To confront economic tht

expla in in a superficia l w

peopie so that they seem to

S O C IAL S C IE N C E AN D N ATU R AL S C IE N C E 247are not peculiar only to s'cientific investigation but are the mode ofordi'ary da ily apprehension of sociar fa -cts. These "priori.tic prin-ciples and the deducrions from them "." "n;il;;,it o,rty by rheprofessional econom-ist but by everybody *rr" a""r, with economicfacts or probrems' The layman dou, ,rot proceed in a way significantlydiferent from thar of trre ,"i",rtirtf ";t;T ; ;*iri"J'i, less criticar,less scrupulous in examining *u""y step in the chain of his deductionsand therefore sometim", -oi" ,,,bj""t to error. O ne need only observeany discussion on current economic problerns to rearize that its courseturns very soon towards a consideration of ubrtru.if"irr"ipt", withoutany reference to experience. you cannot, for instance, discuss thesoviet system withoui fa ting bu"k or rhe genera l principres both ofcapita lisrn and socia ii.m. yo., cannot discuss a wage and hours biilwithout fa lling back on the theory of wages, profits, interests andprices, that means the genera l th"ory of u mu"k"t society. The .,pure

fs61"-lg1 us ser aside the epistemiogica l question whether rhere issuch a thing-is open to different interpretations" These interpretationsrequire e lucidation by theoretica l insight.E conomics is not only not derived f"o*

"op"rience, it is even impos-sible to verify its theorem. tr y upp"ur to "*p""i"nce. E very experienceof a complex phenomenon, we must repea t, can be and is expra inedin different r^/ays' The same facts, th" ,um" .t*ti.ii"ut figures arecla irned as' confirmations of contradictory theories.It is instructive to compare the technique of dearing with experiencein the socia l sciences with that in the natura l ."i"r""r. T 'e have manybooks on econonics which, a fter having developed a theory, annexchapters in which an attempt is made to verify ih" th"ory crevelopedby an appeal to the facts. This i. ,"i *" *"y;hl"t;1"'r"**, scientisrtakes' H e starrs from facrs

"*p"ri**ta lry establiJ"l ;; bu'ds uphis theorv in using them. If rtr *""ty'"'ows a d"dr;;";;a t predictsa state of a ffa irs not yet discovered ir "*p"ri*ents he describes rvhatkind of experiment wourd be cruciar for his theory; the theory seemsto be verified if the result conforms to the p.ediction. This is something

,T *T i# .sisnificantly differen, f*^ the approaJ1"f.", by the

To confront economic theory with rearity we do not have to try toexpla in in a superficia l way " facts in-terpreted differently by otherpeopie so that they seem to verify our theory. This dubious procedu""

248 JO U R N AL O F S O C IAL P H Ii,O S O P H Y AN D JU R IS P R U D E I\C E

is not the way in which reasonable discussion can take place. W hatwe have to do is this: we have to inquire whether the specia l conditionsof action which we have implied in our reasoning correspond to thosewe find in the segment of rea lity under consideration. A theory ofmoney (or ra ther of indirect exchange) is correct or not witlout re fer-ence to the question of whether the actual economic system underexamination employs indirect exchange or only barter.

The method appiied in these theoretica l aprioristic considerationsis the method of speculative constructions. The economist-and like-wise the lay.man in his economic reasoning-buiids up an image ofa non-existent sta te of things. The materia l for this construction isdrawn from an insight into the conditions o{ human action. W hetherthe state of a ffa irs which these speculative constructions depict cor-responds or could correspond to rea lity is irre levant for the ir instru-menta l e fficiency. E ven unrealizable constructions can render va luableservice in giving us the opportunity to conceive what makes them un-rea iizable and in what respect they differ from reatrity. The speculativeconstruction of a socia list community is indispensah'le for economicreasoning notwithstanding the question of whether such a society couldor could not be rea lized.

O ne of the trest known and most frequently applierJ speculativeconstructions is that of a state of static equilibrium mentioned above.W e are fuliy aware that this state can never be rea lized. But we cannotstudy the implications of changes lvithout considering a changelessworld. N o modern economist wili deny that the application of thisspeculative concept has rendered invaluable service in e lucidating thecharacter of entrepreneur's profits and losses and the re la tion betweencosts and prices.

All our economic reasoning operates with these speculative concepts.It is true that the method has its dangers; it easily lends itse lf to errors.But we have to use it because it is the only method available . O f course,we have to be very careful in using it.

To the obvious question, how a pure ly logica l deduction from apri-oristic principles can te ll us any-thing about rea lity, we have to replythat both human thought and human action stem {rom the same rootin that they are both products of the human mind. C orrect results fromour aprioristic reasoning are therefore not only logica lly irre futable ,but a t the same time applicable with a ll the ir apodictic to rea lity

S O C IAL S C IE N

provided that the assumptioway to ref,use a conclusiorconta ins a logica l fa llacy. l

obta ined apply to rea lity. Thstration that the assumptionspart in the rea lity which rvr

The re la tion hetween hiexperience is historica l in tlthing past-and economic tle ra liy assumed. E conomic tlon the contrary the indispens

E conomic history can neitleconomic theory. It is on thtpossible for us to conceive 1

But to orient ourselves inmore lhan mere ly conceireacting man and tire pure lyce ive the categolies of act

besides to understand (uersi

This understanding of thr

of histor:ica l research. The

as possible by the use of a llsciences of human action-1part, economics-and by ti{arther. I{e has to study thcase in question. India iduuthe historian, it is exactly tlor traced back to other entitThe purpose of specific ur

disciplines is to grasp the nprocess. It establishes the I

fixes the va luations, the a inin a word, the tota l philosopwhich they envisaged the c

puts us into the milieu of

S O C IAL S C IE I{C E AN D N ATU R AL S C IE N C E 24gprovided that the assumptions invorved are given in rearity. The onryway to ref,use a concrusion of economics is to demonstrate that itconra ins a logicar fa ilacy. It is another guestion ;;;, rhe resurtsobta ined apply to reariry. This again can be decided only by the demon_stration that the assumptions invorr.ed have or do not tr*" orry counter-F rart in the rea lity rvhich rve wish to expla in.The re la tion hetween historica i experiengs-for every economicexperience is hisrorica l in rhe sense that it is the;;o;;;""" of some.thing pasr-snd economic the 'ry is, there{ore ditr"r;;; ;;om ilrar gen_era liy assumed' E conomic thuory is nct derived fr";;;;"r.ience. * ison the contrary the indispensable iooi for the grorp of """i.*r" history.E conornic history- can neither prove ,ro, dirprove the teachings ofeconornic theory' It is on the-

"oritrury economic theory which makes itpossible for us to conceive th" ".onotic facts "f *," p"ri.

ryBut to orient ourselves in the worrd of hurnan actions we need to domore than merery conceive the meaning of human action. Both theacting man and tlie pure iv obr"rvi"g historian h";; ;;;';nly to con_ce ive the categ'ries of action u. Jororric theory does; they havebesides to undersrand (uersterzuol ,i"-*"aning of human choice .This understanding of the -""ri"g of actioi i- arr"-.o""rfic methodof histor:ica l research. The histori"n'hu. to estabrish the facts as faras possible by rhe ,r".:f a ' the means provided both by the theoretica lsciences of, human

p a r r, e c o n o-, "._ ;:n T ;fJT ,l:fi ;,$ j: ril|;

:?_f ;J i::"1 ;r;3farther' F {e has to study the indiviJ,rut *rra unique conditions of, thecase in quesrion. Ind,ia id.uurn est ineffabile . Individuality is given tothe historian' it is exactry that which'""rrnot be exhaustivery expla inedor traced back to other entities. rn this sense individuarity is irra tionar.The purpose of specific unders,""di"; as applied by the historicardisciplines is to grasp rhe meanin* "i-lla i"ia""rrrirrr"" iu1-"rrotog;""rprocess' It estabiishes the fact th"at we face .Jilrhril #Jiuia ,rut. t,fixes the va luations, the a ims, thu th"oii"., the be lie fs

"rJ ,rr" errors,in a word, the tota l philosophy "r trr"l.ti'g individual, ;;; the way inwhich they envisaged the

"orra itiorr. "ri", which they had to act. Itpurs us into the milieu of the action. or ""*." ,nll *."0c under-

250 JO U R }IAL O F S O C IAL P H ILO S O P H Y AN D JU R IS P R U D E N C B

standing cannot be separated from the phiiosophy of the interpreter.That degree of scientific olijectivity which can be reached in the natura lsciences and in the aprioristic sciences of logic and praxeology can

never be atta ined by the rnora l or historica l sciences (C eistesnissen-

schaften) in the fie ld of the specific understanding. Y ou can under-stand in different ways. H istory can be written from different points

of vierv. The historians may agree in everything that can be establishedin a ra tional way and neverthe less widely disagree in the ir interpre-ta tions. F listory therefore has a lways to he rewritten. l\ew philosophiesdemand a nerr representation of the past.

The specific unclerstanding.of the historica l sciences is not an act ofpure ra tionality. It is the recognition that reason has exhausted a llits resources and that we can do nothing more than to try as well as

we may to give an explanation of something irra tional which is resist-

ant to exhaustive and unique description. These are the tasks which theunderstanding has to fulfill. It is, notwithstanding, a logica l tool andshould be used as such. It should never he abused for the purpose ofsmuggling into the historica l work obscuranticisrn, rnysticism and simi-la r e lements. It is not a free charter for nonsense.

It is necessary to emphasize this point because it sometirnes happens

that the abuses of a certa in type of historicism are justified by an appealto a wrongly interpreted "understanding." The reasoning of logic,praxeology and of the natura l sciences can under no circumstancesbe invalidated by the understanding. H owever strong the evidence sup-

plied by the historica l sources may be, and however understandablea fact may be from the point of view of theories contemporaneouswith it, if it does not fit into our ra tionale , we cannot accept it. Theexistence of witches and the practice of witchcraft are abundantlyattested by legal proceedings; yet we will not accept it. Judgments ofmany tribunals are on record asserting that people have depreciated a

country's currency by upsetting the balance of payments; yet we willnot be lieve that such actions have such effects.

It is not the task of history to reproduce the past. An attempt to doso would be va in and would require a duplication not humanly possible .H istory is a representation of the past in terms of concepts. Thespecific concepts of historica l research are type concepts. These typesof the historica l method can be built up only by the use of the specificunderstanding and they are meaningful only in the frame of the under-

S O C IAL S C IE I

standing to which they owe

concept which is logica llypurpose o{ understanding.if a il the e lements united irfeature . C iasses do not exisl

the mind which in observingIt is another question whetland based on sound conside

given data . There is for inste

which includes not only ItaS panish system of G enera lH orthy and some other syst

way and that it can be contr

includes the R ussian Bols

H ungary and of the short S

classification and the inferlast twenty years divided inis the right way to understa

to question. Y ou can undt

different way by using oth,

and Tota litarianism, and 1

W estern C apita iist system

Boishevism and what the o

you apply the first or the

mode in which you see thJ

classification to be used" a

standing.The type-concepts of the

schaften) are not statisticaclassification are not subjecl

renders it impossible to co

type-concepts (in G erman ,

distinguish them from the

of the triologica l ones) ougl

concepts used for the conc

For instance: The concept'to signify a specific functifuture . In this respect ever

S O C IAL S C IE N C E AN D N ATU R AL S C IE N C E

standing to which they owe the ir existence. Therefore not every type-concept which is logica lly va lid can be considered as useful for thepurpose of understanding. A classification is va lid in a logica l senseif a il the e lements united in one class are characterized by a commonfeature . C lasses do not exist in actuality, they are a lways a product ofthe mind which in observing things discovers likenesses and differences.It is another question whether a classification which is logica lly va lidand based on sound considerations can be used for the expianation ofgiven data . There is for instance no doubt that a type or class'"Fascism"which includes not only Ita iian Fascism but a lso G erman N azism, thespanish system of, G enera l F ranco, the H ungarian system of Admira lH orthy and some otler systems can be constructed in a logica lly va lidway and that it can be contrasted to a type ca lled 'oBolshevism," whiclrincludes the R ussian Bolshevism and the system of Bela Kun inH ungary and of the short soviet episode o{ M unich. But whether thisclassification and the inference from it which sees the worid of thelast twenty years divided into the two parties, Fascists and Bolsheviks,is the right way to understand present-day politica l conditions is opento question. Y ou can understand this period of history in a quitedifferent way by using other types. Y ou may distinguish D emocracyand rota litarianism, and then le t the type D emocracy include theW estern capita iist system and the type Tota litarianisrn include bothBolshevism and what the other classification terms Fascism. W hetheryou apply ttrie first or the second typification depends on the wholemode in which you see things. The understanding decides upon theciassification to be used, and not the classification upon the under-standing.

The type-concepts of the historica l or mora l sciences (G eistesnissen-schaften) are not sta tistica l averages. M ost of the features used forclassification are not subject to numerica l determination, and this a lonerenders it impossible to construct them as statistica l averages. Thesetype-concepts (in G e'man one uses the term klea l-Typus in order todistinguish them from the type-concepts of othe i sciences, especia llyof the biologica l ones) ought not to be confused with the praxeologica lconcepts used for the conceiving of t}e categories of human action.For i'stance: The concept o'entrepreneur" is used in economic theoryto signify a specific function, that is the provision for an uncerta infuture . In this respect everybody has to some extent to be considered

25L

Y

252 JO U R N AL O F S O C IAL P H ILO S O P H Y AN D JU R IS P R U D E N C E

as an entrepreneur. O f course, it'is not the task of this classification

in economic theory to distinguish men, but to distinguish functions and

to expla in sources of profit or loss. E ntrepreneur in this sense is the

personification of the function which results in profit or loss. In

""onoroi" history and in dealing with current economic problems the

term o'entrepreneur" signifies a class of men wlio are engaged in busi-

ness but who may in many other respects differ so much tha i the

genera l term entrepreneul seems to be meaningless and is used only

rvith a specia l qualification, for instance big (medium'sized, smali)

business, "W -a ll S treetr" a lmaments business, G erman business, e tc. f

The type entrepreneur as used in history and politics can never have I

the conceptual exactitude which the praxeologica i concept entrepreneur l

has. Y ou never meet in life men who are nothing e lse than the personifi-

cation of one function onlY .2

V

The preceding remarks justify the conclusion that there is a radica l

difference between the methods of the socia l sciences and those of the

natura l sciences. The socia l sciences owe the ir plogress to the use of

the ir particular methods and have to go further a long the iines which

the specia l character of the ir object require . They do not have to adopt

the rnethods of the natura l sciences.

It is a fa llacy to recommend to the socia l sciences the use of mathe '

matics ancl to be lieve that they could in this way be made more "exact."

The application of mathematics does not render physics more exact

or more certa in. Let us quote E inste in's rernark: 'oAs far as rnathe '

matica l propositions re fer to rea lity they are not certa in and as faras they are certa in they do not re fer to rea lity." It is different with

praxeologica l propositions. These re fer with a ll the ir exactitude and

certa inty to the rea lity o{ human action. The explanation of this phe-

nomenon lies in the fact that both-the science of human action and

human action itse lf-have a cornmon roa \ i.e ., human reason. It would

be a mistake to assume that the quantita tive approach could render

them more exact. E very numerica l expression is inexact because of the

inherent limita tions of human powers of measurement. For the rest we

2 For the sake of completeness we have to remark that there is a third use of the term

"ttt"pt"""otin law which has to be carefully distinguished from the two mentioned above '

S O C IAL S C IE

have to re fer to what has

character of quantita tive etThe re formers who wish

the methods of the natur,e fforts by pointing to the Ideny that the socia l scienc

being perfect. E very econoBut two considerations mulsatisfactory state of socia lwith an a lleged inadequaclthe teachings of economicsblame the discipline for thday be necessary to re formnot take its direction a longThe objections of these are

N ational Bwreau of E conon

?l

I

S O C IAL S C IE N C E AN D N ATU R AL S C IE N C E 253

have to re fer to *:: ti:]::1 ::li-",i"## ii:r:"ffil.rTH '"""11

character of quantita tive expresslons rn

Thereforrnerswhowishtoimprovethesocia lsciencesbyadoptingthemethodsofthenatura lsciencessometimestrytojustifythe irefforts by pointing to the backward state of the former' N obody will

deny that the socia l sciences and especia lly economics are far from

being perfe"t. E u"'y """""*f" krrows.how much remains to be done '

But two "orrside 'a tions

must be kept in mind' F irst: The present un'

satisfactory state of socia l and economic conditions has nothing to do

with an a lleged inadequacy in economic theory' If people do not use

the teaching, "i ""o"o*it' u' a guide for the ir policies they cannot

blamethedisciplineforthe irownfa ilure 'secondly:Ifitmaysomedaybenecessalytoreformeconomictheoryradica llythischangewillnot take its direction a long the lines suggested by the present critics'

itt" ol;""tions of these are thoroughly re futed forever'

N ational Bwreau of E conomic R esearch

t

1

II