Modul

e3

TheBehavi
our
alRevol
uti
onandt
heSci
ent
ifi
cQuest

I
nt he 1 950s and t he 1 960s,a behavi our alr evol ution st ress ing t he
systemat i
c and r i
gor ousl y s cient i
f i
c st udy of i ndi vidualand gr oup
behavi oursweptpol i
t i
calsci encedi sci pli
ne.Af ocusonst udyi ngpol it
ical
behavi our ,r ather t han i nstit
ut i
ons or i nt erpret at i
on of l egalt ext s,
charact er ized ear ly behavi our alpol iti
calsci ence,i ncl udi ng wor k by
Rober tDahl ,Davi dEast onandPhi l
li
pConver se.
Poli
ticalbehavi ourr ef erst oapar ticul arappr oach,as etofmet hodsf or
thest udyofpol iti
cs ,andt hest udyofhumanbehavi ouri npol iti
cs.Letme
pointoutt hatbehavi our alism isnotanewf i
el dofPol iti
calSci ence;r at her
i
ti s a new appr oach t ot he st udy of pol iticaland gover nment al
phenomena.The goalofbehavi our alism i st o make pol iticalsci ence a
mor es cient if
icdi scipl i
ne,onewhi chanal ys espol iti
csempi rically,t hati s,
usi
ngt hesci ent i
ficmet hod.
Thebehavi ouralappr oachadopt edbybehavi oural istsisdi stinguishedby
i
tsat temptt odescr ibegover nmentasapr ocessmadeupoft heact i
ons
andi nt eract ionsofi ndi vidualsandgr oupsofi ndi vidual s.I ti sconcer ned
witht heact ivi
tiesofgover nment s,pol iti
calpar ties ,inter es tgr oupsand
voter s.Unl ike inst itutionalt heor et icalappr oach,behavi our alappr oach
rejectedi nstit
utionsast heuni tofanal ys i
sbecause,gover nment ,af ter
all
,isnotmadeupmer elyofdocument s,cont ainingl awsandr ul
esorof
str
uct ur esofapar ti
cul arf orm buti sf undament allybaseduponpat ter ns
of act i
ons and act iviti
es of men.Thus,t he uni t s of anal ysis of
behavi our al i
sm shoul d be t he obs er ved behavi ourand r elat i
ons hipsof
i
ndividual s.
Beginni ngf rom 1 950s,behavi our allyi nclinedpol iti
calsci ent istsbor rowed
thenat uralsci ent i
st s’appr oachandaccumul ateds t
at i
sticsf rom el ect ions ,
publi
c opi ni on sur veys,vot es i nl egi slat ures,and anyt hing el s et hey
could hang a numberon.Speci fical ly,t he t erm " behavi our alsci ence" ,
subsequent ly cal l
ed " behavi our alism" , was coi ned by a gr oup of
quant i
tat ivel y or ient ed,r i
gorous l
y i nclined s oci als cient i
sts,at t he
Univer sityofChi cago.

TheQuestt obeSci enti

fi
c
Aspolit
icalanal ysts,weendeavourt obe‘ obj ecti
veinouranal ysi
s,sothatour
l
ivesordi sli
kes,bi ases,val
ues,orper sonalities’donotdi st
ortouranalysis.We
alsowantt oensur ethatourst udiesprovideexpl anati
onsandgener al
izati
ons
whichcanbet estedorver i
fi
edbyot heranal ysts.Bysodoi ng,weseekt obe
“sci
enti
fic”,i
.e.tobesyst ematicinourst udies .Thi
sisthesensei nwhichwear e
Poli
ti
calScientists,i.
e.weseekt ouset hesci entifi
cmet hod.
Youmayhaveal wayst houghtt hat“sci
ence’i sexclusi
velyusedinr el
ati
ont o
thenat uralsci encesl i
keChemi s t
r y,Biology andPhysi cs.Thi sisnott hecas e
because sci ence -f rom t he Lat in wor d‘ scire’-si mply means ‘ to know” ,
consequent ly,anyst udywhi chadvancesknowl edgei sscient i
fi
c.Today,mos t
peopl et hi
nkofsci enceaspr eciseandf actual,suppor t
edbyexper i
ment sand
data.Butt hisi snotal l
.Sci encecoul dbeamat terofmet hod,i.e.how t ost udy,
orofsubst ance,i .
e.whati sst udied,orbot h.Thenat uralsciencesar esci ent i
fi
c
i
n bot h senses.Pol iti
calSci ence l ike the ot hersoci alsciences ( Sociology,
Economi cs,Psychol ogy,HumanGeogr aphy)i ssci entifi
conl yint hesenseofi ts
met hodsbecausei tseekst oappl yt hesyst emat i
ct oolsofst udy ofmani n
society.Some pol i
ticalanal ysts have at tempt ed t o become l ike nat ural
scientists;t hey quant if
y dat a and mani pulatet hem st ati
sticall
yt o val idate
hypot heses. Al though t hese pol it
ical scientists have made some good
cont ri
but i
ons,t hey f ocuson s mal lquest i
onsofdet ailrathert han on l arge
quest i
onsofmeani ng.Youshoul dbei nformedt hatt hi
sisasar esultofl imited
areasofpol i
ticalsciencet hatcanbequant ifi
ed:publ i
copi nion,vot ert urnout ,
electionr etur nsandl egislat i
vevot i
ng.
Opinionsar ehoweverdi videdonwhet herpol i
ti
cscanbest udi edinasci ent i
fi
c
mannersuchaswoul ddi sall
ow ourown pr ejudi cesorval uest o colourour
analysi s.I
nt henextl ect ur e,Iwillexami net hear gument sagai nstasci enceof
polit
ics.Butj ustnow,Iwantt odi scusst heessent i
alelement soft hesci ent i
fi
c
met hod,andhowpol iti
calanal ystshavesoughtt obesci enti
fic.

AnOutli
neofScienti
ficMet hod
Ishal
lbegi
nbypr esenti
nginbr oadout
li
ne,t
hebasi
cpr
ocedur
eoft
hesci
ent
ifi
c
method.Asimpl
eout l
ineinvolves:

1.Makingexpl ici
t( i
n acl earerf orm),in advance,yourassumpt ionsand
expect ationsaboutwhatyouar estudying;
2.Makingexpl i
cit,i
nadvance,t her ul
esand/orgui del
inesbywhi chyouwi l
l
proceedi nyourst udy.Youdonotchanget herulesofaf oot ballgameat
half-
time.
3.Making car efulobser vat i
ons of t he phenomena i n whi ch you ar e
i
nterest ed,wi t
havi ew t odiscoveringtheel ementsofthest udyandt he
regulariti
es( consistentpat ternsofbehavi ourofoccur r
ence)whi chmay
exist.
4.Seeking,t hrough one ormor e expli
ci
tf r amesofr ef
erence ( theor
ies,
perspect i
ve)t o mapoutt he relati
onshipsamongt he t
hingsyouhave
obser ved;
5.Seekingt oexpl ainther elati
onshipsyoui dent i
fy.
6.Makingver i
fiable(t
estable)pr edi
cti
onsbasedonyourexpl anat i
onsand
7.Repor tingf ullyandcl early,yourconcl usionsi nsuchawayt hatanot her
Scientist coul d,i f he wi shed,r epeat t he study by maki ng si milar
obser vations.

I
sitpossi
blet
ohaveasci
ent
ifi
cst
udyofpol
i
tics?
Ai
msofSci
enceofPol
i
tics
ASci
enceofpol
i
ticsai
msatt
hef
oll
owi
ngi
mpor
tantgoal
s:
1.Val ue- Fr ee Anal ysi s: Thi sr efer st o t he quest f or obj ect i
vi ty and
neut ralityi npol i
ticalanal ys is.T obesci ent ifi
c,t heanal ystmustanal yze
facts( dat a)ast heyar er athert hanort heyoughtt o/shoul dbe.Asmuch
aspossi bl e,ourper sonall ikesand di sl i
kes,pr efer ences,i nter est sor
valuesorbi asesmustbekeptoutofouranal ysis.
2.Empi ricalanal ysis:concer nwi t h‘whati s’r athert han‘ whatoughtt obe’
impliesan emphasi son di rectobs ervat ion t o di scovert hingsast hey
reall
yar e,t heirr elat i
onshi pwi t
hot hert hings,andt her egulariti
esoft hei r
occur r ence.I tison obser vedr egular i
tiesoft hei roccur rence.I ti son
these obser ved r egularities t hat we pr emi se our expl anat ions and
predict ions.
3.Expl anat i
on: sci ent i
fic expl anat ions appeal t o gener ali
zat i
on and
theor iesi nexpl ainingspeci ficoccur rence.I ft hesegener ali
zat i
onsand
thepar t
icul arcondi tions/pr emi sesar et rue,t hent heconcl usion( s)mus t
bet rue.
4.Pr edi ct i
on:t akest hesamel ogi calf or m asexpl anat i
on,buti sdi f ferent
because i t i s f or war d l ooki ng, and i nvol ves speci fyi ng
condi tions/pr emi sesunderwhi chcer tainoccur r
encesar el i
kel yt ot ake
place.
5.Theor ies:asci ent ifictheor yi sas etofgener ali
zat i
onwhi chspeci fiest he
direction ofr el
at ionshi psamong var i
abl es.Theor i
esar et her ef oret he
maj ori ngr edient sofexpl anat i
on.Butf ort hem t ober eall
yhel pf uli nt hi s
regar d,t hey shoul d be gener aland r est ricti
ve.Fi nal ly,agood t heor y
shoul dbeopent of ur t
herempi ricaltes ts.
6.Laws: ar est atement sofuni ver saluni for mitieswhi chr elat et oal lt he
casesofapar t
icul arphenomenoni .e.,t hey donotal low f orexcept ion.
They ar e usef ulf orbot h expl anat i
ons and pr edict i
ons,butdo not
possessasmuchexpl anat or y powerast heor iesdo t hought hey have
great ercer t
ainty.
I
saSci enceofPol i
ti
csdesi rable?
Thedesi rability ofsci enceofpol i
ti
cshasbeendet er mi nedi nt hel astl ect ur e.
Youwer et oldt hatt hei mpor tanceofsci enceofpol iti
csort heuseofsci ent ifi
c
met hodi npol iticalanal ysi sliesi nt hef actt hati twi llhel pt oexpl ainandt o
predictont hebasi sofr egular i
tieswhi chhavebeenobser vedandf oundt obe
rel
iable.Itisdesi rablet ousesci ent i
ficmet hodi nor dert ohaveananal ysi st hat
i
s val ue –f ree,empi ri
caland ver i
fiablet hat wi l
lpr oduce gener ali
zat i
ons ,
theoriesandl awst hatcanbeus edt oexpl ainandpr edict .Thenextr elevant
question rel atingt ot heuseofsci ent i
fi
cmet hodi n pol i
ticalanal ysi si s“ i
sa
scienceofpol iti
cspossi bl e?
I
saScienceofPol i
ti
csPossibl
e?
Nowwear einapositi
ont oanswert hequest
iononthepossi
bil
i
tyofscienceof
pol
it
ics orpoli
ti
calanalysisthatmakes use ofscienti
fi
c method.We shall
howeverrelyont hewr i
ti
ngsofnumer ousauthor
swhobel i
evet hatascience
ofpoli
ti
csisnotpossi
ble.Weshallt
aket hemajori
ssuesoneaftertheother:
1
.ValueFr eeAnal ysis:Somepol iti
calscienti
stsbelievethatiti
sdiffi
cult,i
f
notimpossi ble,f orpolit
icalanalysist o be value-f
ree.You may have
cert
ain per sonalr easons for deci di
ng t o study localgover nment
administrati
onr atherthansay,el ectoralbehaviour.Youmaybel i
evef or
exampl et hatel ecti
onsar e notf ree and fair
,and so,do notr equi
re
anal
ysi s.Once youhave sel ect ed yourt opi
c,i tbecomesdi f
f i
cultfor
yourval uest o beel i
minatedf r
om youranal ysisbecauseyouar ean
i
nterestedpar tofwhatyous tudy.Thisiswhy younor mal
lyf i
ndt hat
peopleanal ysisoft hesameeventdi ffer,somet i
mes so mar kedly,that
youf i
ndi tdiffi
culttobelievethattheyar eanalyzi ngthesamet hi
ng.
The otherprobl
ems i s,isi tdesi rabl
ef ort he pol
it
icalanal
ystt o be
objecti
veforitsownsake?Af terknowingt hi
ngsast heyare(assuming
thatheisvalue-
free),shouldt hepol i
ti
calanalystnotgoaheadt ot el
lus
what ought to be? Wi thout doubt ,pol i
ti
cal sci
entist
s do have a
responsi
bi
li
tytosoci etyast heyar einvolvedi nthesearchforabet t
er
societ
y.Wouldtheirquestf orobj ecti
vit
yf oritsownsakenotr educethe
rel
evanceofpolit
icalscientists?

2.Empir i
calanal ysi s: Thisr estson“ har d”fact swhi char eobser vableand
capableofbei ngsubj ectedt ol abor ator
yandquant i
t ati
veanal ysis.But ,
canwer eall
ygetsuch“ hard”f actinpol i
tics?Muchofwhatwest udyi s
man’sbehavi our ,wecannotr elyonwhatwet hi
nkmadehi m behavei na
parti
cularway.Wehavet omost ly dependonwhathet el
lsusandt hi
s
may notber eliablebecaus eman i scapabl eofl ying.Thi sisdi fferent
from thehar df actsi nsayphysi csorchemi strywhi chcanbedescr i
bed
i
npur el
y physi calt er msbasedonobser vat ion.I npol i
tics,evensucha
simple act i
on l i
ke vot ing cannotbe descr i
bed as a pur ely physi cal
acti
vity.
I
f direct obser vat i
on and har df acts ar e diff i
cultt hen quant i
tative
analysisi smor e di ffi
cult.First,we cannotsubj ectmen t ot he same
l
abor atory condi ti
ons underwhi ch nat uralsci ent i
sts car ry outt heir
analysis.Asar esul t,ifwer eall
yseekt obesci ent i
f i
c,wewoul dhavet o
concent rateonpol i
ticalphenomenawhi chcanbedi rectlyobser vedand
are quant i
fiable.Thi s diver ts attent i
on away f rom t he mai n stuf fof
poli
ti
csl ikel eader ship,anddeci sionmaki ngwhi chcannotbesubj ecting
poli
ti
calphenomena t o empi ricalanal ysis.Opi nion pol ls and sur vey
research met hods whi ch ar e bas ed on quest i
onnai res have been
conduct ed,and comput ers and advanced st atisti
cal( mat hemat i
cal)
techniqueshavebeendevi sedtost udypol i
ticalphenomena.

3.Uncertai
nti
esandUnpr edi
ctabi
lit
yinhumanl i
fe:Theot heressenti
alof
thescienti
fi
cmet hod-explanationandpr edi
cti
onswhi char ebasedon
theor
iesandl aws-may bes ummar i
zedby sayingthatt hey al
lrel
y on
observedregular
it
iesinparti
cularoccur
rences.I
not herwor ds,t
heyr ely
onconsistentpatter
nsofoccur r
encestobeabl etoexplainandpr edict
.
 Thi
si swher e a science ofpol i
ti
cs ispar t
icularl
y handicapped.Man’s
behaviourr emainsuncer tainandunpredictable,nomat t
erhow muchwe
know hi m about .Consequent l
y,itisdiffi
culttof ormulat
euniversalor
generalt heor i
es much l ess“ l
aws” because t here would al
ways be
exceptionst oobservedr egulari
ti
es.Aslongast hiscannotbeover come,
ourexpl anati
onsandpr edicti
onswil
lremainincompl eteandinadequat
e.
Some scholarscr it
ici
ze a science ofpol i
ti
cson t he basisoft he f
oll
owing
fact
ors:
(
a)The Compl exit
y of polit
icalphenomena:one argument against t
he
possibi
li
ty of science of poli
ti
cs cl
aims t
hatno regul
ari
ti
es can be
discovered because poli
ti
calphenomenon aretoo complex withtoo
manyvar i
ablesandpossiblerel
ati
onshi
ps.
(
b)Human indeterminancy:Human behavi
ori s sai
dt o be unpr
edict
abl
e.
Russel
lKirk,putsitthi
sway “Humanbei ngsar etheleastcontr
oll
abl
e,
veri
fi
able,l
awobeyi ngandpredi
ctabl
eofsubjects.
(
c)TheReact i
onpr oblem:Someoft hecr i
ti
csofscient
ifi
cpoli
ti
calscience
pointoutthatsincet hesubjectsorr espondent
sareawar eoft hefact
thatthey arebei ngstudi
ed,theirresponsescannotbet aken asval i
d
i
ndicator
soft heiropini
ons.
(
d)The I
nfl
uence ofval
ue:Unl
i
ke t
he soci
alscient
ist
s,practi
ti
oners of
natur
alsci
encesdonothavetodealwit
hvalues– biases,pr
ejudi
ces,
l
ikesanddisl
i
kes.
(
e)The change phenomenon:Change affect
s ourconcl
usi
on on human
behaviour because human bei
ngs ar
e dynamic and change i
sthe
constantfactorofli
fe.
I
ti simpor t
antt oi nform you thatmuch ofpol i
ti
csespecially when one i s
deali
ng with how and why deci si
ons are made i sj usttoo compl ex t o be
quantif
ied or measur ed.Polit
ical sci
ence is an empi r
icaldi sci
pli
ne t hat
accumul at
esbot hquant i
fiedandqual i
tati
vedat a.Inspiteoft hesedi f
ficul
ti
es,
west il
lhavegener al
izati
onsandt heor i
eswhi ch,tot heextentt hatthey ar e
stat
edempi r
ical
ly andt heref
oreopent oconf ir
mat i
onori nformat i
on,canbe
consideredscienti
fic.Ifwecannotbepur el
y scienti
fi
c,atleastwecant ryt o
approximatescience.
TheDebat
eCont
inues
Thedebat eoverwhet heras ci
enceofpol i
ti
csisposs i
bleordes i
rableisnotlikelyt obe
completelyresolved.Thetradi
tionali
sts,i.
e.thosepol i
ti
calsci
ent i
stswhoar guet hatiti
s
notpos si
bleto besci enti
fi
candemphasi zenor mativetheoriescer t
ainly havet heir
poi
nts,asdo t he empiri
ci
stsorbehavi oural
is
ts( s
ee nextl ect ure)who s eek to be
sci
entif
ic.Atthepr esents t
age,t heagr eements eemst obet hatpol i
ti
calphenomena
canbes t
udiedwhi l
ethosewhi chcannotbes ubjectedtoscient i
ficanalysi
sbes tudi
ed
phi
losophical
lyus i
ngnormat i
vet heori
es.
Meanwhi
le,poli
ti
calsciencealsotakesafteranat
uralsci
encewhentheresearcher
s,i
f
theyar
epr ofessi
onal,studythingsastheyareandnotast heywishthem tobe.You
needtoknow thatmostpol it
icalsci
ent
ist
shaveviewpoint
soncurr
entissues,anditi
s
notdiff
icul
tt olettheseviewst ai
ntt hei
ranalysesofpoli
ti
cs.Ishoul
dl etyouknow
thatany scienti
fi
canal ysi
sofpol i
ti
csr equi
resfourmajoringredi
ents.A sci
enti
fi
c
analysi
sofpol i
ti
cali
ssuesors chol
arlywor k“shoul
dbereasoned,bal
anced,suppor
ted
byevidence,andabi ttheoret
ical
.”
Reasoned analyt
icalwor kr equi
resyout o spelloutyourr easoning,and itshoul
d
makes ense.I
tisimper at
ivef oryout osay so,ifyourperspecti
vei scoloredby an
underl
yingassumption.Forinstance,youmights ay,‘
forthepurposeoft hisanalysi
s,
weas sumethatpeopl earecoher entorrat
ional,
’or‘Thi
sisaneval uati
veanal ysi
sof
femalevoter
sinar uralarea.
’
Anotherwaybywhi chyoucanmi ni
mizebiasisbyacknowl edgi
ngt hatthereareother
waysofl ooki
ngatyourt opi cori ssueatst ake.Thi
siswhati scal l
ed abal anced
analysi
s.Forabalancedanalysis,youneedt oment i
onvarioustheoreti
calapproaches
thatyouconsideredrel
evanti nt hel
iterat
uretoyourt opi
candwhatt heyst andf or
.
Youmay go f urt
herto cr
it
icizet hemai n ar
gumentofeachoft heappr oachesand
explai
nwhyyout hi
nktheyar eweak,i nadequateorfaul
ty.
Also,youranal ysismustbesuppor tedwi t
hevi dence.Itisamaj orr equi
rementf oral
l
scholarl
y or s ci
enti
fi
c analysisto be s upported with quant i
fiable or quali
tat
ive
evidenceorbot h.Inthecas eofpoli
ti
calanalys i
s,itut
il
izesbot h.Apar tf
rom common
knowl edgewhi chdoesnothavet obesuppor t
ed( forexampl e,youdonotneedt o
i
nterview thePr esi
dentofNi ger
iato‘prove’thatNigeri
aobt ainedheri ndependencei n
st
October1 1 960)anys tatementthatisexposedt ocontrover syori nter
pretat
ionhad
betters upport
edwi thevidence.Ifi
tishardt of i
ndempi r
icalevidence,atami ni
mum,
yououghtt ociteas cholarwhohasmuchevi dencet odemonst rateori l
l
ustrat
et he
pointyouar etryi
ngtoexpr ess.
Lastl
y,at hought
f ulanal
ysisi
sal waysconnected,atleastalitt
le,to atheoreti
cal
pointofview.Whileyoumaynotneedasweepi ngnew theor
y,itshouldadvancet he
di
scipl
ine’
sknowl edge.Iti
sexpect edthatyourtheory atami ni
mum,shoul deither
confir
m orr ef
uteanexi sti
ngtheory.Meredescri
pti
onofs omethi
ngi snotat heory,
whichiswhyWi ki
pediaorGoogl
ear enotconsi
deredadequateforpoli
ti
calanal
ys i
s.