The British Society for the Philosophy of Science

Historical Explanation in the Social Sciences

Author(s): J. W. N. Watkins
Source: The British Journal for the Philosophy of Science, Vol. 8, No. 30 (Aug., 1957), pp.
104-117
Published by: Oxford University Press on behalf of The British Society for the Philosophy of
Science
Stable URL: http://www.jstor.org/stable/685790 .
Accessed: 10/06/2014 12:17

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Oxford University Press and The British Society for the Philosophy of Science are collaborating with JSTOR to
digitize, preserve and extend access to The British Journal for the Philosophy of Science.

http://www.jstor.org

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICALEXPLANATION IN THE
SOCIAL SCIENCES*
J. W. N. WATKINS

I Introduction
THE hope which originally inspired methodology was the hope of
finding a method of enquiry which would be both necessary and
sufficient to guide the scientist unerringly to truth. This hope has
died a natural death. Today, methodology has the more modest task
of establishing certain rules and requirements which are necessary to
prohibit some wrong-headed moves but insufficient to guarantee
success. These rules and requirements, which circumscribe scientific
enquiries without steering them in any specific direction, are of the
two main kinds, formal and material. So far as I can see, the formal
rules of scientific method (which comprise both logical rules and
certain realistic and fruitful stipulations) are equally applicable to all
the empirical sciences. You cannot, for example, deduce a universal
law from a finite number of observations whether you are a physicist,
a biologist, or an anthropologist. Again, a single comprehensive
explanation of a whole range of phenomena is preferable to isolated
explanations of each of those phenomena, whatever your field of
enquiry. I shall therefore confine myself to the more disputable (I had
nearly said ' more disreputable') and metaphysically impregnated part
of methodology which tries to establish the appropriate material
requirements which the contents of the premisses of an explanatory
theory in a particular field ought to satisfy. These requirements may
be called regulative principles. Fundamental differences in the subject-
matters of different sciences-differences to which formal methodolo-
gical rules are impervious-ought, presumably, to be reflected in the
regulative principles appropriate to each science. It is here that the
student of the methods of the social sciences may be expected to have
something distinctive to say.
* A revisedversionof a paperread at the FirstAnnual Conferenceof the
Philosophyof ScienceGroup,Manchester,on 23rd September1956. Footnotes
havebeenaddedsubsequently.
Io04

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES
An example of a regulativeprincipleis mechanism,a metaphysical
theory which governed thinking in the physical sciences from the
seventeenthcentury until it was largely supersededby a wave or field
world-view. According to mechanism, the ultimate constituentsof
the physical world are impenetrable particles which obey simple
mechanicallaws. The existenceof these particlescannot be explained
-at any rate by science. On the other hand, every complex physical
thing or event is the resultof a particularconfigurationof particlesand
can be explained in terms of the laws governing their behaviour in
conjunction with a description of their relative positions, masses,
momenta, etc. There may be what might be describedas unfinished
or half-way explanationsof large-scalephenomena (say, the pressure
inside a gas-container)in termsof other large-scalefactors(the volume
and temperatureof the gas); but we shall not have arrivedat rock-
bottom explanations of such large-scale phenomena until we have
deduced their behaviour from statements about the properties and
relationsof particles.
This is a typically metaphysicalidea (by which I intend nothing
derogatory). True, it is confirmed, even massively confirmed, by
the huge successof mechanicaltheorieswhich conform to its require-
ments. On the other hand, it is untestable. No experiment could
overthrow it. If certain phenomena-say, electromagnetic pheno-
mena-seem refractory to this mechanisticsort of explanation, this
refractorinesscan always (and perhaps rightly) be attributedto our
inability to find a successfulmechanicalmodel ratherthan to an error
in our metaphysicalintuition about the ultimate constitution of the
physicalworld. But while mechanismis weak enoughto be compatible
with any observationwhatever,while it is an untestableand unempirical
principle,it is strong enough to be incompatiblewith variousconceiv-
able physical theories. It is this which makes it a regulative,non-vacuous
metaphysicalprinciple. If it were compatiblewith everythingit would
regulate nothing. Some people complain that regulative principles
discourage researchin certain directions, but that is a part of their
purpose. You cannot encourage researchin one direction without
discouragingresearchin rival directions.
I am not an advocate of mechanism but I have mentioned it
becauseI am an advocate of an analogous principlein social science,
the principle of methodological individualism.' According to this
1 Both of these analogous principles go back at least to Epicurus. In recent
times methodological individualism has been powerfully defended by Professor
H los

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
principle, the ultimate constituentsof the social world are individual
people who act more or less appropriatelyin the light of their dis-
positions and understandingof their situation. Every complex social
situation,institution,or event is the resultofa particularconfiguration
of individuals, their dispositions, situations, beliefs, and physical
resourcesand environment. There may be unfinished or half-way
explanationsof large-scalesocial phenomena (say, inflation) in terms
of other large-scalephenomena (say, full employment); but we shall
not have arived at rock-bottom explanations of such large-scale
phenomenauntil we have deducedan accountof them from statements
about the dispositions,beliefs, resources,and inter-relationsof indivi-
duals. (The individuals may remain anonymous and only typical
dispositions,etc., may be attributedto them.) Andjust as mechanism
is contrastedwith the organicistidea of physicalfields, so methodolo-
gicalindividualismis contrastedwith sociologicalholism or organicism.
On this latter view, social systems constitute 'wholes ' at least in the
sense that some of their large-scalebehaviouris governed by macro-
laws which are essentiallysociologicalin the sense that they are sui
generis and not to be explained as mere regularitiesor tendencies
resulting from the behaviour of interacting individuals. On the
contrary,the behaviourof individualsshould (accordingto sociological
holism) be explainedat least partly in terms of such laws (perhapsin
conjunctionwith an account, first of individuals'r6les within institu-
tions and secondly of the functions of institutionswithin the whole
social system). If methodological individualism means that human
beings are supposedto be the only moving agents in history, and if
sociologicalholism meansthat some superhumanagents or factorsare
supposed to be at work in history, then these two alternativesare
exhaustive. An example of such a superhuman,sociological factor is
the allegedlong-term cyclicalwave in economic life which is supposed
to be self-propelling, uncontrollable, and inexplicable in terms of
human activity, but in terms of the fluctuationsof which such large-
scale phenomena as wars, revolutions, and mass emigration,and such
F. A. Hayek in his IndividualismandEconomic Orderand The Counter-Revolution
of Science,and by ProfessorK. R. Popperin his The OpenSocietyandits Enemies
and ' The Povertyof Historicism' Economica,1944-45,11-12. Followingin their
footstepsI have also attemptedto defendmethodologicalindividualismin 'Ideal
Types and Historical Explanation' this Journal, 1952, 3, 22, reprinted in Readings
in thePhilosophy
of Science,ed. FeiglandBrodbeck,New York, 1953. This article
hascome in for a good dealof criticism,thechiefitemsof whichI shalltryto rebut
in whatfollows.
I06

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES

psychologicalfactorsas scientificand technologicalinventivenesscan,

it is claimed, be explained and predicted.
I say 'and predicted' because the irreducible sociological laws
postulated by holists are usually regardedby them as laws of social
development, as laws governing the dynamics of a society. This
makes holism well-nigh equivalent to historicism, to the idea that a
society is impelled along a pre-determinedroute by historical laws
which cannotbe resistedbut which can be discernedby the sociologist.
The holist-historicistposition has, in my view, been irretrievably
damagedby Popper'sattackson it. I shall criticisethis position only
in so far as this will help me to elucidateand defend the individualistic
alternativeto it. The centralassumptionof the individualisticposition
-an assumption which is admittedly counter-factual and meta-
physical-is that no social tendency exists which could not be altered
if the individualsconcernedboth wanted to alter it and possessedthe
appropriateinformation. (They might want to alter the tendency
but, through ignoranceof the factsand/or failureto work out some of
the implicationsof their action, fail to alterit, or perhapseven intensify
it.) This assumptioncould also be expressedby saying that no social
tendency is somehow imposed on human beings 'from above' (or
'from below ')-social tendenciesare the product (usuallyundesigned)
of human characteristicsand activities and situations, of people's
ignorance and laziness as well as of their knowledge and ambition.
(An example of a social tendency is the tendency of industrialunits to
grow larger. I do not call 'social' those tendencieswhich are deter-
mined by uncontrollablephysicalfactors,such as the alleged tendency
for more male babies to be born in times of diseaseor war.) 1
1 The issue of holism versusindividualismin socialsciencehas recentlybeen
presentedas though it were a questionof the existenceor non-existenceof irre-
duciblysocialfactsratherthanofirreduciblysociologicallaws. (SeeM. Mandelbaum,
'Societal Facts', TheBritish Journalof Sociology,1955,6, and E. A. Gellner,'Ex-
planationsin History', AristotelianSociety,SupplementaryVolume 30, 1956.)
This way of presentingthe issueseemsto me to empty it of most of its interest.
If a new kindof beastis discovered,whatwe wantto knowis not so muchwhether
it falls outsideexisting zoologicalcategories,but how it behaves. People who
insiston the existenceof socialfactsbutwho do not saywhetherthey are governed
by sociologicallaws, arelike peoplewho claimto have discoveredan unclassified
kind of animalbut who do not tell us whetherit is tameor dangerous,whetherit
can be domesticatedor is unmanageable.If an answerto the questionof social
factscouldthrow light on the seriousandinterestingquestionof sociologicallaws,
then the questionof socialfactswould also be seriousandinteresting. But this is
not so. On the one hand,a holistmayreadilyadmit(asI pointedout in my' Ideal
I0o7

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
My procedure will be: first, to de-limit the sphere in which
methodological indivisualism works in two directions; secondly, to
clear methodological individualism of certain misunderstandings;
thirdly, to indicate how fruitful and surprising individualistic explana-
tions can be and how individualistic social theories can lead to sociolo-
gical discoveries; and fourthly, to consider in somewhat more detail
how, according to methodological individualism, we should frame
explanations, first for social regularities or repeatable processes, and
secondly for unique historical constellations of events.

2 WhereMethodological Does not Work

Individualism
There are two areasin which methodological individualismdoes
not work.
The first is a probabilitysituationwhere accidentaland unpredict-
able irregularitiesin human behaviour have a fairly regular and
predictable overall result?. Suppose I successively place 1,000 indi-
viduals facing north in the centre of a symmetricalroom with two
exits, one east, the other west. If about 500 leave by one exit and
about 500ooby the other I would not try to explainthis in terms of tiny
undetectable west-inclining and east-inclining differences in the
individuals,for the samereasonthat Popperwould not try to explain
Types' paper,which Gellnercriticises)thatall observablesocialfactsarereducible
to individualfactsandyet hold thatthe latterareinvisiblygovernedby irreducibly
sociologicallaws. On theotherhand,anindividualist mayreadilyadmit(asGellner
himselfsays)thatsomelargesocialfactsaresimplytoo complexfor a full reduction
of themto be feasible,andyet hold that individualistic explanations of themarein
principlepossible,just as a physicistmay readilyadmit that some physicalfacts
(forinstance,thepreciseblast-effectsof a bomb-explosion in a built-uparea)arejust
too complexfor accuratepredictionor explanationof them to be feasibleandyet
hold thatpreciseexplanations andpredictionsof themin termsof existingscientific
lawsarein principlepossible.
This revisedway of presentingthe holism versusindividualismissuedoes not
only divertattentionfromthe importantquestion. It alsotendsto turnthe dispute
into a purelyverbalissue. ThusMandelbaum is ableto provethe existenceof what
he calls 'societal facts' becausehe definespsychologicalfacts very narrowlyas
'factsconcerningthethoughtsandactionsof specifichumanbeings' (op.cit. p. 307).
Consequently, of anonymous
thedispositions individuals whichplaysuchanimportant
r61ein individualisticexplanationsin socialscienceare 'societal facts' merelyby
definition.
1 Failureto exclude probability-situations from the ambit of methodological
individualism wasanimportantdefectof my' IdealTypes' paper. Here,Gellner's
criticism(op. cit. p. 163)doeshit the nailon the head.
Io8

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES
the fact that about 500oo balls will topple over to the west and about
500ooto the east, if balls are dropped from immediately above a
1,000ooo
north-south blade, in terms of tiny undetectablewest-inclining and
east-inclining differences in the balls. For in both cases such an
'explanation' would merely raise the furtherproblem: why should
these west-inclining and east-inclining differences be distributed
approximatelyequallyamong the individualsand among the balls ?
Those statisticalregularitiesin social life which are inexplicablein
individualisticterms for the sort of reasonI have quoted here are, in a
sense,inhuman,the outcome of a largenumberof sheeraccidents. The
outcome of a largenumberof decisionsis usuallymuch less regularand
predictablebecausevariablehumanfactors(changesof taste,new ideas,
swings from optimism to pessimism)which have little or no influence
on accident-ratesare influential here. Thus Stock Exchange prices
fluctuate widely from year to year, whereas the number of road-
accidents does not fluctuate widely. But the existence of these
actuarialregularitiesdoes not, as has often been alleged, support the
historicistidea that defencelessindividualslike you and me are at the
chance mercy of the inhuman and uncontrollabletendencies of our
society. It does not supporta secularisedversion of the Calvinistidea
of an Almighty Providence who picks people at random to fill His
fixed damnation-quota. For we can control these statisticalregular-
ities in so far as we can alter the conditions on which they depend.
For example, we could obviously abolish road-accidentsif we were
preparedto prohibit motor-traffic.
The second kind of social phenomenon to which methodological
individualism is inapplicable is where some kind of physicalconnection
between people'snervoussystemsshort-circuitstheirintelligentcontrol
and causesautomatic, and perhapsin some sense appropriate,bodily
responses. I think that a man may more or less literally smell danger
and instinctivelybackaway from unseenambushers; andindividuality
seems to be temporarily submerged beneath a collective physical
rapportat jive-sessions and revivalist meetings and among panicking
crowds. But I do not think that these spasmodicmob-organismslend
much support to holism or constitute a very serious exception to
methodological individualism. They have a fleeting existence which
ends when their members put on their mufflersand catch the bus or
otherwise disperse,whereasholists have conceived of a social whole as
something which enduresthrough generationsof men; and whatever
holds together typical long-lived institutions,like a bank or a legal
Iog

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
system or a church, it certainlyis not the physicalproximity of their
members.

3 Misunderstandings Individualism
of Methodological
I will now clear methodological individualism of two rather
widespreadmisunderstandings.
It has been objected that in making individual dispositions and
beliefs and situationsthe terminusof an explanationin social science,
methodological individualism implies that a person's psychological
make-up is, so to speak, God-given, whereasit is in fact conditioned
by, and ought to be explained in terms of, his social inheritanceand
environment.1 Now methodologicalindividualismcertainlydoes not
prohibit attemptsto explain the formation of psychologicalcharacter-
istics;it only requiresthat such explanationsshould in turn be indivi-
dualistic,explainingthe formation as the result of a seriesof conscious
or unconsciousresponsesby an individual to his changing situation.
For example, I have heard Professor Paul Sweezey, the Harvard
economist, explainthat he became a Marxistbecausehis father,a Wall
Streetbroker, sent him in the I930's to the London School of Econo-
mics to study underthose staunchliberaleconomists,ProfessorsHayek
and Robbins. This explanationis perfectlycompatiblewith methodo-
logical individualism(thoughhardlycompatible,I shouldhavethought,
with the Marxist idea that ideologies reflect class-positions) because it
interprets his ideological development as a human response to his
situation. It is, I suppose, psycho-analystswho have most system-
atically worked the idea of a thorough individualist and historical
explanation of the formation of dispositions, unconscious fears and
beliefs, and subsequentdefence-mechanisms,in terms of responsesto
emotionally charged,and especiallychildhood, situations.
My point could be put by sayingthat methodologicalindividualism
encouragesinnocentexplanationsbut forbidssinisterexplanationsof the
widespreadexistence of a dispositionamong the members of a social
group. Let me illustratethis by quoting from a reply I made to
Goldstein'scriticisms.
1 Thus Gellner writes : 'The real oddity of the reductionist [i.e. the methodolo-
gical individualist's]case is that it seems to preclude a priorithe possibility of human
dispositionsbeing the dependent variablein an historicalexplanation-when in fact
they often or always are ' (op. cit. p. I65). And Mr Leon J. Goldstein says that in
making human dispositions methodologically primary I ignore their cultural
conditioning. (TheJournalof Philosophy,I956, 53, 807.)
IIO

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES

Suppose that it is establishedthat Huguenot traderswere relatively

prosperous in I7th-century France and that this is explained in terms
of a wide-spread disposition among them (a disposition for which
there is independent evidence) to plough back into their businessesa
larger proportion of their profits than was customary among their
Catholic competitors. Now this explanatory disposition might very
well be explained in its turn-perhaps in terms of the general thrifti-
ness which Calvinism is said to encourage, and/or in terms of the
fewer alternative outlets for the cash resources of people whose re-
ligious disabilities prevented them from buying landed estates or
political offices. (I cannot vouch for the historical accuracy of this
example.)
I agree that methodological individualism allows the formation,
or ' culturalconditioning ', of a widespreaddispositionto be explained
only in terms of other human factors and not in terms of something
inhuman,such as an alleged historicist law which impels people willy-
nilly along some pre-determined course. But this is just the anti-
historicistpoint of methodological individualism.
Unfortunately, it is typically a part of the programme of Marxist
and other historicist sociologies to try to account for the formation of
ideologies and other psychological characteristicsin strictly sociological
and non-psychological terms. Marx, for instance, professed to believe
that feudal ideas and bourgeoisideas are more or less literally generated
by the water-mill and the steam-engine. But no description, however
complete, of the productive apparatus of a society, or of any other
non-psychological factors, will enable you to deduce a single psycho-
logical conclusion from it, because psychological statements logically
cannot be deduced from wholly non-psychological statements. Thus
whereas the mechanistic idea that explanations in physics cannot go
behind the impenetrable particles is a prejudice (though a very under-
standable prejudice), the analogous idea that an explanation which
begins by imputing some social phenomenon to human factors cannot
go on to explain those factors in terms of some inhuman determinant
of them is a necessary truth. That the human mind develops under
various influences the methodological individualist does not, of course,
deny. He only insists that such development must be explained
'innocently' as a series of responses by the individual to situations and
not 'sinisterly' and illogically as a direct causal outcome of non-
psychological factors, whether these are neurological factors, or
impersonal sociological factors alleged to be at work in history.
Another cause of complaint against methodological individualism
III

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
is that it has been confused with a narrow species of itself (Popper calls
it 'psychologism') and even, on occasion, with a still narrower
sub-species of this (Popper calls it the' Conspiracy Theory of Society ').1
Psychologism says that all large-scale social characteristicsare not
merely the intended or unintendedresultof, but a reflectionof, indivi-
dualcharacteristics.2ThusPlato saidthatthe characterandmake-upof
a polis is a reflectionof the characterand make-up of the kind of soul
predominant in it. The conspiracy theory says that all large-scale
social phenomena (do not merely reflect individual characteristics
but) are deliberately brought about by individuals or groups of
individuals.
Now there are social phenomena,like massunemployment,which
it would not have been in anyone's interest deliberately to bring about
and which do not appearto be large-scalesocialreflectionsor magnified
duplicatesof some individualcharacteristic. The practicalor techno-
logical or therapeuticimportance of social science largely consists in
explaining, and thereby perhapsrenderingpolitically manageable,the
unintended and unfortunateconsequencesof the behaviour of inter-
1 See K. R. Popper, The Open Societyandits Enemies,2nd edn., 1952, ch. 14
2 I am at a loss to understandhow Gellner came to make the following strange
assertion: '... Popper refers to both " psychologism " which he condemns,
and " methodological individualism ", which he commends. When in the articles
discussed[i.e., my" IdealTypes "paper] " methodological individualism"is worked
out more fully than is the case in Popper's book, it seems to me to be indistinguish-
able from "Psychologism".' Finding no difference between methodological
individualism and a caricature of methodological individualism, Gellner has no
difficultyin poking fun at the whole idea : ' Certain tribesI know have what anthro-
pologists call a segmentary patrilineal structure,which moreover maintains itself
very well over time. I could " explain " this by saying that the tribesmen have,
all or most of them, dispositions whose effect is to maintain the system. But, of
course, not only have they never given the matter much thought, but it also might
very well be impossible to isolate anything in the charactersand conduct of the
individual tribesmen which explains how they come to maintain the system'
(op. cit. p. 176). Yet this example actually suggests the lines along which an
individualistic explanation might be found. The very fact that the tribesmen
have nevergiven the mattermuchthought,the fact that they accept their inherited
system uncritically,may constitutean importantpartofan explanationof its stability.
The explanationmight go on to pin-point certainrules-that is firm and widespread
dispositions-about marriage, inheritance,etc., which help to regularisethe tribes-
men's behaviour towards their kinsmen. How they come to share these common
dispositions could also be explained individualisticallyin the same sort of way that
I can explain why my young children are already developing a typically English
attitude towards policemen.
II2

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES

acting individuals. From this pragmaticpoint of view, psychologism

and the conspiracytheory are unrewardingdoctrines. Psychologism
says that only a change of heart can put a stop to, for example, war
(I think that this is BertrandRussell'sview). The conspiracytheory,
faced with a big bad social event, leadsto a hunt for scape-goats. But
methodological individualism, by imputing unwanted social pheno-
mena to individuals'responsesto their situations,in the light of their
dispositions and beliefs, suggests that we may be able to make the
phenomena disappear,not by recruiting good men to fill the posts
hitherto occupied by bad men, nor by trying to destroy men's socially
unfortunate dispositionswhile fostering their socially beneficial dis-
positions,but simply by alteringthe situationsthey confront. To give
a currentexample, by confrontingindividualswith dearermoney and
reduced credit the Government may (I do not say will) succeed in
halting inflationwithout requiringa new self-denying attitude on the.
part of consumersand without sending anyone to prison.

4 FactualDiscoveriesin SocialScience

To explain the unintendedbut beneficial consequencesof individual

activities--by 'beneficial I
consequences' mean social consequences
which the individuals affected would endorse if they were called on to
choose between their continuationor discontinuation--isusuallya task
of less practical urgency than the explanation of undesirableconse-
quences. On the other hand, this task may be of greatertheoretical
interest. I say this because people who are painfully aware of the
existence of unwanted social phenomena may be oblivious of the
unintended but beneficial consequencesof men's actions, rather as a
man may be oblivious of the good health td which the smooth
functioning of his digestion,nervous system, circulation,etc., give rise.
Here, an explanatorysocial theory may surpriseand enlighten us not
only with regardto the connectionsbetween causesand effectbut with
regardto the existence of the effect itself. By showing that a certain
economic system contains positive feed-back leading to increasingly
violent oscillations and crises an economist may explain a range of
well-advertised phenomena which have long been the subject of
strenuous political agitation. But the economists who first showed
that a certain kind of economic system contains negative feed-back
which tends to iron out disturbancesand restoreequilibrium,not only
II3

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
explained, but also revealed the existence of, phenomena which had
hardly been remarkedupon before.'
I will speak of organic-like social behaviour where members of
some social system (that is, a collection of people whose activities
disturb and influence each other) mutually adjust themselves to the
situationscreatedby the othersin a way which, without directionfrom
above, conduces to the equilibriumor preservationor development
of the system. (These are again evaluativenotions, but they can also
be given a' would-be-endorsed-if' definition.) Now such far-flung
organic-likebehaviour,involving people widely separatedin spaceand
largelyignorantof each other, cannotbe simply observed. It can only
be theoreticallyreconstructed--by deducing the distant social conse-
quencesof the typicalresponsesof a largenumberof interactingpeople
to certain repetitive situations. This explains why individualistic-
minded economistsand anthropologists,who deny that societiesreally
are organisms, have succeeded in piecing together a good deal of
unsuspected organic-like social behaviour, from an examination of
individualdispositionsand situations,whereassociologicalholists, who
insist that societies really are organisms,have been noticeably unsuc-
cessfulin convincingly displayingany organic-likesocial behaviour-
they cannot observeit and they do not try to reconstructit individual-
istically.
There is a parallel between holism and psychologism which
explainstheircommon failureto makesurprisingdiscoveries. A large-
scale social characteristicshould be explained, according to psycholo-
gism, as the manifestation of analogous small-scale psychological
tendenciesin individuals,and accordingto holism as the manifestation
of a large-scale tendency in the social whole. In both cases, the
explicansdoes little more than duplicatethe explicandrum.The method-
ological individualist,on the other hand, will try to explain the large-
1 This sentence, as I have since learnt from Dr A. W. Phillips, is unduly compla-
cent, for it is very doubtful whether an economist can ever show that an economic
system containing negative feed-back will be stable. For negative feed-back may
produce either a tendency towards equilibrium, or increasing oscillations, accord-
ing to the numerical values of the parametersof the system. But numericalvalues
arejust what economic measurements,which are usuallyordinalratherthan cardinal,
seldom yield. The belief that a system which contains negative feed-back, but
whose variablescannot be describedquantitatively,is stable may be based on faith or
experience, but it cannot be shown mathematically. See A. W. Phillips,' Stabilis-
ation Policy and the Time-Forms of Lagged Responses', The EconomicJournal,
1957, 67.
II4

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES
scale effect as the indirect,unexpected, complex product of individual
factors none of which, singly, may bear any resemblanceto it at all.
To use hackneyedexamples,he may show that a longing for peace led,
in a certain internationalsituation, to war, or that a government's
desire to improve a bad economic situation by balancing its budget
only worsened the situation. Since Mandeville'sFableof the Beeswas
published in 1714, individualisticsocial science, with its emphasison
unintendedconsequences,has largely been a sophisticatedelaboration
on the simple theme that, in certainsituations,selfishprivate motives
may have good social consequencesand good political intentions bad
social consequences?.
Holists draw comfort from the example of biology, but I think
that the parallelis really between the biologist and the methodological
individualist.The biologist does not, I take it, explain the large
changeswhich occur during,say, pregnancy,in termsof corresponding
large teleological tendenciesin the organism, but physically, in terms
of small chemical, cellular,neurological,etc., changes,none of which
bearsany resemblanceto theirjoint and seeminglyplanfuloutcome.

5 How SocialExplanationsShouldbe Framed

I will now considerhow regularitiesin social life, such as the trade
cycle, should be explainedaccordingto methodologicalindividualism.
The explanationshould be in terms of individualsand their situations;
and since the process to be explained is repeatable,liable to recur at
various times and in various parts of the world, it follows that only
very general assumptionsabout human dispositionscan be employed
in its explanation. It is no use looking to abnormalpsychology for an
explanation of the structure of interest-rates-everyday experience
1 A good dealof unmeritedoppositionto methodological individualismseems
to springfrom the recognitionof the undoubtedfact that individualsoften run
into socialobstacles. Thus the conclusionat which Mandelbaumarrivesis 'that
therearesocietalfactswhich exerciseexternalconstraintsoverindividuals' (op. cit.
p. 317). This conclusionis perfectlyharmoniouswith the methodologicalindi-
vidualist'sinsistencethatplansoftenmiscarry(andthatevenwhen they do succeed,
they almost invariablyhave other importantand unanticipatedeffects). The
methodologicalindividualistonly insiststhatthe socialenvironmentby which any
particularindividualis confrontedand frustratedand sometimesmanipulatedand
occasionallydestroyedis, if we ignore its physicalingredients,made up of other
people,theirhabits,inertia,loyalties,rivalries,andso on. Whatthe methodological
individualistdeniesis thatan individualis everfrustrated, or destroyed
manipulated
or bornealongby irreduciblesociologicalor historicallaws.
IIS

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 J. w. N. WATKINS
must contain the raw materialfor the dispositional(as opposed to the
situational) assumptionsrequired by such an explanation. It may
requirea strokeof genius to detect, isolate,and formulatepreciselythe
dispositionalpremissesof an explanationof a social regularity. These
premissesmay state what no one had noticed before, or give a sharp
articulationto what had hitherto been loosely described. But once
stated they will seem obvious enough. It took years of groping by
brilliantmindsbefore a preciseformulationwas found for the principle
of diminishingmarginalutility. But once stated,the principle-that
the less, relatively, a man has of one divisible commodity the more
compensationhe will be disposed to require for foregoing a small
fixed amount of it-is a principle to which pretty well everyone will
give his consent. Yet this simple and almost platitudinousprinciple
is the magic key to the economics of distributionand exchange.
The social scientistis, here, in a position analogousto that of the
Cartesianmechanist.1 The latter never set out to discover new and
unheard-of physical principles because he believed that his own
principle of action-by-contact was self-evidently ultimate. His
problem was to discover the typical physical configurations, the
mechanisms,which, operatingaccordingto this principle,produce the
observedregularitiesof nature. His theoriestook the form of models
which exhibited such regularitiesas the outcome of'self-evident'
physical principlesoperating in some hypotheticalphysical situation.
Similarly, the social scientist does not make daring innovations in
psychology but relies on familiar,almost 'self-evident' psychological
material. His skill consists,first in spotting the relevant dispositions,
and secondly in inventing a simple but realisticmodel which shows
how, in a precise type of situation, those dispositionsgenerate some
typical regularityor process. (His model, by the way, will also show
that in this situation certain things cannot happen. His negative
predictionsof the form, ' If you've got this you can'thave that as well '
may be of great practicalimportance.) The social scientistcan now
explainin principlehistoricalexamplesof this regularprocess,provided
his model does in fact fit the historicalsituation.
This view of the explanationof socialregularitiesincidentallyclears
up the old questionon which so much ink hasbeen spilt about whether
the so-called 'laws' of economics apply universally or only to a
particular'stage' of economic development. The simple answer is
that the economic principlesdisplayedby economists' models apply
I owe this analogy to Professor Popper
116

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions
 HISTORICAL EXPLANATION IN SOCIAL SCIENCES

only to those situationswhich correspondwith their models; but a

single model may very well correspondwith a very large number of
historicalsituationswidely separatedin space and time.
In the explanation of regularitiesthe same situationalscheme or
model is used to reconstructa number of historicalsituationswith a
similarstructurein a way which revealshow typical dispositionsand
beliefsof anonymousindividualsgenerated,on each occasion,the same
regularity.1 In the explanationof a unique constellationof events the
individualisticmethod is again to reconstructthe historicalsituation,
or connected sequence of situations, in a way which reveals how
(usuallyboth named and anonymous) individuals,with their beliefs
and dispositions(which may include peculiarpersonaldispositionsas
well as typical human dispositions), generated, in this particular
situation, the joint product to be explained. I emphasisedispositions,
which are open and law-like, as opposed to decisions,which are occur-
rences, for this reason. A person's set of dispositionsought, under
varying conditions, to give rise to appropriatelyvarying descisions.
The subsequent occurrence of an appropriate decision will both
confirm, and be explainedby, the existenceof the dispositions. Sup-
pose that a historical explanation (of, say, the growth of the early
Catholic Church) largely relies on a particular decision (say, the
decision of Emperor Constantine to give Pope Silvester extensive
temporalrightsin Italy). The explanationis, so far, ratherad hoc: an
apparently arbitraryfiat plays a key r61e in it. But if this decision can
in turnbe explainedas the offspringofa marriageofa set of dispositions
(for instance,the Emperor'sdispositionto subordinateall rival power
to himself) to a set of circumstances(for instance, the Emperor's
recognitionthat Christianitycould not be crushedbut could be tamed
if it became the official religion of the Empire), and if the existence of
these dispositions and circumstancesis convincingly supported by
independentevidence, then the area of the arbitrarilygiven, of sheer
brute fact in history, although it can never be made to vanish, will have
been significantlyreduced.

The LondonSchoolof EconomicsandPoliticalScience

The Universityof London

1 This should rebut Gellner'sconclusionthat methodologicalindividualism

would transformsocialscientistsinto ' biographers
engrandeserie' (op. cit. p. 176).

117

This content downloaded from 195.78.109.194 on Tue, 10 Jun 2014 12:17:18 PM

All use subject to JSTOR Terms and Conditions