On he Vi ues and Vices o Combining Theo ies:
The Case o Ins i u ional and Ac o -Ne wo k Theo ies in Accoun ing Resea ch
S en Modella,b,c,*, Eija Vinna id and Ka i Lukkac
aAlliance Manches e Business School,Uni e si y o Manches e , Uni ed Kingdom
b NHH—No wegian School o Economics, Be gen, No way
c Tu ku School o Economics, Tu ku, Finland
d Uni e si y o Tampe e, Finland
Abs ac
This pape examines he possibili ies and challenges o combining me hod heo ies in
accoun ing esea ch h ough an analysis o s udies which combine insigh s om ins i u ional
heo y (IT) and ac o -ne wo k heo y (ANT). We in es iga e he pa adigma ic challenges
associa ed wi h combining hese me hod heo ies and whe he and how schola s ha e deal
wi h such challenges. We demons a e how he combina ion o hese me hod heo ies in a single
s udy gi es ise o conside able pa adigma ic ensions. The mos signi ican ensions ela e o
he wo me hod heo ies’ di e ging on ological concep ions o he na u e o social s uc u es
and agency and hei e y di e en epis emological iews o he ole o heo y. Mo eo e , ou
e iew o ex an accoun ing esea ch combining IT and ANT indica es ha a la ge numbe o
s udies simply igno e such ensions and do no p o ide deepe e lec ions on he pa adigma ic
implica ions o combining hese me hod heo ies. Whils ecognizing he subs an i e
con ibu ions eme ging om his body o esea ch, we ques ion whe he con inued
app ochemen be ween IT and ANT is he mos app op ia e way o wa d and sugges
al e na i e heo e ical pa hs o examining he ins i u ionaliza ion o accoun ing. We also call
on accoun ing esea che s o exe cise much g ea e e lexi i y ega ding he pa adigma ic
implica ions o combining me hod heo ies as well as he mo e gene al jus i iabili y o such
p ac ices as a ehicle o ad ancing ou unde s anding o accoun ing as a social and
o ganiza ional p ac ice.
Key wo ds: accoun ing, ac o -ne wo k heo y, ins i u ional heo y, pa adigms, heo e ical
plu alism.
In oduc ion
O e he pas h ee decades, accoun ing esea ch e ol ing wi hin he in e -disciplina y, o
“al e na i e”, adi ion has o med a ib an esea ch p og amme held oge he by a s ong
commi men o heo e ical plu alism (see e.g., Lukka & Mou i sen, 2002; Bax e & Chua,
2003; Llewellyn, 2003; Pa ke , 2012). The main idea o such plu alism is o allow, and e en
encou age, he use o a b oad ange o me hod heo ies (Lukka & Vinna i, 2014) a he han
con ining he choice o hose me hod heo ies de i ed om economics o psychology as is
ypical o mains eam accoun ing esea ch.
1
Whils his commi men o heo e ical plu alism
1
Lukka & Vinna i (2014, p. 1309) de ine domain heo y as “a pa icula se o knowledge on a subs an i e opic
a ea si ua ed in a ield o domain such as accoun ing” and me hod heo y as “a me a-le el concep ual sys em o
s udying he subs an i e issue(s) o he domain heo y a hand”. In addi ion o economics and psychology,
accoun ing schola s ha e applied me hod heo ies o igina ing om, o ins ance, o ganiza ion s udies and
This is an accep ed manusc ip . The o iginal a icle has been published in Accoun ing,
O ganiza ions and Socie y. 2017, ol. 60, pp. 62-78. DOI:h p://dx.doi.o g/10.1016/
j.aos.2017.06.005.
2
mani es s i sel in a a ie y o ways, an issue a ac ing inc easing a en ion is he p opensi y
o accoun ing schola s o combine di e se me hod heo ies in a single s udy and he
oppo uni ies and challenges ha his c ea es (e.g., Co aleski, E ans, Lu & Shields, 2003;
Jacobs, 2012; Hoque, Co aleski & Goone a ne, 2013; Bea ie, 2014; O’Dwye & Une man,
2016). Such combina ions can ange om he selec i e bo owing and inco po a ion o
elemen s o one me hod heo y wi hin ano he , dominan me hod heo y o ull- ledged
blending o me hod heo ies in an a emp o gene a e “new” heo ies (Oswick, Fleming &
Hanlon, 2011; Suddaby, Ha dy & Huy, 2011). Bo h ypes o combina ions can be an impo an
sou ce o heo e ical eju ena ion and a e ela i ely unp oblema ic as long as he me hod
heo ies being combined a e no oo a apa in e ms o on ological and epis emological
assump ions (Gioia & Pi e, 1990; Schul z & Ha ch, 1996; Kakku i-Knuu ila, Lukka &
Kuo ikoski, 2008; Maye & Spa owe, 2013).
Howe e , in ecen yea s, conce ns ha e inc easingly been aised abou he endency o
esea che s o combine me hod heo ies wi h e y di e en , and e en incompa ible, on ologies
and epis emologies
2
and he challenges his b ings o he ask o econciling con lic ing
assump ions in he p ocess o heo y de elopmen . Such conce ns ha e eme ged in in e -
disciplina y accoun ing esea ch (e.g., Modell, 2013, 2015a, 2015b) as well as he b oade
managemen and o ganiza ion li e a u e (e.g., Okhuysen & Bona di, 2011; Thompson, 2011)
and wa an se ious conside a ion ega dless o how ex ensi e he blending o me hod heo ies
is.
3
Whils he combina ion o me hod heo ies wi h incompa ible on ologies and
epis emologies does no necessa ily in alida e he subs an i e insigh s eme ging om such
esea ch, i equi es a e lexi e app oach o heo y de elopmen . Acco ding o Okhuysen and
Bona di (2011, p. 10), such esea ch needs o be accompanied by “a deep discussion o how
unde lying assump ions can be combined, and especially whe he his combina ion can eally
be achie ed wi hou s aining agains logical impossibili ies”. Howe e , esea che s combining
a ious me hod heo ies do no always heed such ad ice. Fo ins ance, Coope , Ezzamel and
Willmo (2008) and Modell (2015a) show ha esea che s combining ins i u ional and c i ical
heo ies end o p i ilege he o me and, as a esul , comp omise key on ological and
epis emological assump ions o he la e wi hou o e ing deepe e lec ions on such p ac ices.
Recen e iews and deba es ega ding he combina ion o me hod heo ies in managemen
accoun ing (Hoque e al., 2013, 2015; Modell, 2015b) and public sec o accoun ing esea ch
(Jacobs, 2012, 2013; Modell, 2013) sugges ha such endencies owa ds un e lexi e
eclec icism may be mo e widesp ead in he b oade , in e -disciplina y accoun ing esea ch
communi y.
The objec i e o his pape is o ex end he deba e on he pa adigma ic implica ion o combining
me hod heo ies in in e -disciplina y accoun ing esea ch h ough a sys ema ic analysis o
sociology. We employ his pai o concep s o highligh he auxilia y ole ha me hod heo ies play in accoun ing
s udies aiming o con ibu e o di e se domain heo ies.
2
On ology and epis emology a e di icul o de ine b ie ly wi hou doing injus ice o hei complexi y. While
acknowledging his di icul y, in his pape , on ology e e s o di e en heo ies’ assump ions ega ding he na u e
o eali y (Boyd, Gaspe & T ou , 1991; o a ypology o such assump ions, see e.g. Law, 2004, pp. 24-25), whils
epis emology e e s o he na u e o scien i ic knowledge and how such knowledge can be jus i ied (Boyd e al.,
1991; Rosenbe g, 2016).
3
Acco ding o Okhuysen and Bona di (2011), he ela i e a en ion o on ological and epis emological issues is
likely o a y somewha depending on whe he one dominan heo y bo ows selec i ely om o he heo ies o
whe he he heo ies being combined a e placed on a mo e equal oo ing. Howe e , bo h ypes o combina ions
equi e a highly e lexi e app oach o heo y de elopmen .
3
esea ch which combines insigh s om ins i u ional heo y (IT)
4
and ac o -ne wo k heo y
(ANT) and, in doing so, we d aw a en ion o he “ i ues” and “ ices” o such esea ch
p ac ices. In his seminal discussion o he de elopmen o ins i u ional accoun ing esea ch,
Lounsbu y (2008) iden i ied his pa icula combina ion o me hod heo ies as a po en ially
p omising way o wa d. Whils accoun ing esea ch in o med by IT (see Dilla d, Rigsby &
Goodman, 2004; Ribei o & Scapens, 2006) and ANT (see Jus esen & Mou i sen, 2011; Lukka
& Vinna i, 2014) has e ol ed in o wo subs an ial bodies o li e a u e in hei own igh , we
also iden i y a g owing s eam o esea ch combining insigh s om hese me hod heo ies.
Simila a emp s o combine IT and ANT can be ound in he managemen and o ganiza ion
li e a u e and ha e in o med se e al s eams o esea ch, e ol ing unde he ub ics o
Scandina ian ins i u ionalism (Cza niawska & Se on, 1996; Sahlin & Wedlin, 2008),
ins i u ional wo k (Law ence & Suddaby, 2006; Ra iola & No bäck, 2013) and ins i u ional
logics (Jones, Boxenbaum & An hony, 2013). These de elopmen s may be seen as a p omising
a enue o ad ancing esea ch in o he ins i u ionaliza ion o o ganiza ional p ac ices.
Howe e , as ou analysis shows, he combina ion o IT and ANT cons i u es a ela i ely
ex eme example o me hod heo ies based on incompa ible on ological and epis emological
assump ions. Explo ing such an example allows us o es he limi s o how a he ambi ion o
combine me hod heo ies in con empo a y accoun ing esea ch can be aken and o c i ically
e alua e how esea che s ha e deal (o no deal ) wi h he pa adigma ic ensions ha a ise
om such endea ou s. Ou analysis e eals a widesp ead lack o e lexi i y ega ding key
pa adigma ic ensions associa ed wi h he combina ion o IT and ANT on he pa o
esea che s and leads us o p oblema ize he claims made by Lounsbu y (2008) and o he s ha
he wo me hod heo ies can be use ully combined. Mo e gene ally, we call on accoun ing
esea che s o exe cise much g ea e e lexi i y conce ning he pa adigma ic implica ions o
combining me hod heo ies. This is pa icula ly impo an when he me hod heo ies being
combined es on di e ging on ological and epis emological assump ions. Gi en ha we a e
explo ing a ela i ely ex eme example o such esea ch, ou obse a ions should no be aken
as an a gumen o a hal o he combina ion o me hod heo ies. Howe e , we cau ion agains
he endencies owa ds un e lexi e eclec icism, which a e occasionally associa ed wi h such
p ac ices, and u ge accoun ing schola s o also e lec on wha makes he combina ion o
me hod heo ies a alid schola ly endea ou in a mo e gene al, epis emic sense.
The pape p oceeds as ollows. In he ollowing sec ion, we compa e he on ological and
epis emological assump ions o IT and ANT and discuss he key pa adigma ic ensions which
a e likely o eme ge when he wo me hod heo ies a e combined in a single s udy. We hen
p esen he esul s o ou e iew o ex an accoun ing li e a u e combining IT and ANT. Finally,
we discuss ou indings and hei implica ions o u u e esea ch.
Ins i u ional and Ac o -Ne wo k Theo y: A Compa ison o On ological and
Epis emological Assump ions
The ollowing sec ion compa es he assump ions embedded in he on ological ounda ions and
epis emological commi men s o IT and ANT. Simila o Schul z and Ha ch (1996), we pay
pa icula a en ion o how he e olu ion o hough wi hin each o hese bodies o esea ch has
gi en ise o di e ences and simila i ies in such assump ions and whe he his c ea es
oppo uni ies o combining he wo heo ies in a logically cohe en way. A e discussing each
4
In he p esen pape , we con ine he no ion o IT o new (o neo-) ins i u ional sociology (see G eenwood, Oli e ,
Sahlin & Suddaby, 2008), which has cons i u ed he key ins i u ional app oach o app ochemen wi h ANT in
he accoun ing li e a u e.
4
heo y in some de ail we o e an ini ial assessmen o which pa adigma ic ensions a e likely
o eme ge when IT and ANT a e combined as me hod heo ies in a single s udy and summa ize
he main issues o in e es in ou e iew o he accoun ing li e a u e pu suing such a
combina ion.
Ins i u ional heo y
On ological ounda ions
In e ms o on ology, IT is g ounded in a social cons uc i is iew o he wo ld bu has
unde gone a numbe o changes, implying a a ying deg ee o ideli y o hese o igins, since
i s eme gence in he la e 1970s. In hei ounda ional wo k, Meye and Rowan (1977) d ew
hea ily on Be ge and Luckman (1967) and desc ibed ins i u ions as social cons uc ions which
g adually achie e a aken- o -g an ed, o objec i ied, s a us ha ende s social s uc u es
ela i ely insensi i e o he immedia e in luence o social ac o s. E en hough Meye and
Rowan (1977) ecognized he ole o collec i e human agency in he ep oduc ion and
ans o ma ion o ins i u ions, he emphasis on social s uc u es as ela i ely objec i ied and
immu able en i ies implies ha he on ology unde pinning IT combines a social cons uc i is
iew wi h a mo e o less p onounced, ealis dimension. This ep esen s a o m o “dep h
on ology” acco ding o which social eali ies a e hie a chically s uc u ed and a leas pa ly
independen o indi idual ac o s a any gi en ime (see Leca & Naccache, 2006). The ealis
concep ion o social s uc u es was a guably ei ied as he heo y e ol ed in he 1980s (Meye ,
Boli & Thomas, 1994; Tolbe & Zucke , 1996; Bow ing, 2000) and ins i u ional en i onmen s
we e po ayed as ela i ely s able and homogeneous o ganiza ional ields, o ecognized a eas
o social li e, which a e shaped by p ocesses o ins i u ional isomo phism a he han delibe a e
human agency (DiMaggio & Powell, 1983; Woo en & Ho man, 2008).
S a ing in he la e 1980s, his o e -socialized iew o ac o s was inc easingly challenged
(DiMaggio, 1988; Oli e , 1991; Powell & DiMaggio, 1991), and i has since gi en way o a
ple ho a o wo k ha has sough o esu ec he social cons uc i is on ology o IT and
examine how social ac o s shape ins i u ions (see Ba ilana, Leca & Boxenbaum, 2009;
G eenwood, Rayna d, Kodeih, Micelo a, & Lounsbu y, 2011). This signi ies a ansi ion o an
on ological posi ion ha places g ea e emphasis on he agency implica ed in a ec ing
s uc u al change wi hou nega ing he linge ing and o en signi ican in luence o p e-exis ing
s uc u es on such changes. Ra he han concei ing o human agency as a ee- loa ing and
ela i ely uncons ained phenomenon, mos ins i u ional heo is s now subsc ibe o a
concep ion o agency as ins i u ionally embedded and o en implica ed in a complex in e play
wi h his o ically con ingen s uc u es (Lounsbu y, 2008; Lounsbu y & Boxenbaum, 2013;
Zilbe , 2013). Ye , o e ime, wo dis inc s ands o esea ch, unde pinned by sligh ly
di e ging on ological emphases, ha e come o domina e IT. One o hese s ands has
main ained much o he s uc u alis emphasis, which cha ac e ized ea ly ad ances in IT, whils
he o he is made up o a ange o app oaches which a e uni ed by a mo e ac o -cen ic
on ology.
The p ima y s uc u alis s and o con empo a y IT esea ch is he one e ol ing unde he
ub ic o ins i u ional logics. This app oach e ol ed om he pionee ing wo k o F iedland and
Al o d (1991) and has paid conside able a en ion o how o ganiza ional ields ei he end o
unde go ela i ely comple e shi s be ween dominan logics o e ol e in o mo e he e ogeneous
and agmen ed en i ies as a esul o being s uc u ed by mul iple and o en compe ing logics
(see Tho n on & Ocasio, 2008; Tho n on, Ocasio & Lounsbu y, 2012). Ins i u ional logics a e
de ined as “ he socially cons uc ed, his o ical pa e ns o ma e ial p ac ices, assump ions,
5
alues, belie s, and ules by which indi iduals p oduce and ep oduce hei ma e ial
subsis ence, o ganize ime and space, and p o ide meaning o hei social eali ies” (Tho n on
& Ocasio, 2008, p. 101). Ins i u ional logics ha e a dis inc on ological s a us in ha hey shape
he ways in which indi iduals and g oups o social ac o s concei e o he wo ld. Consis en
wi h he no ion o human agency as an ins i u ionally embedded phenomenon, he concep ion
o agency associa ed wi h ins i u ional change is one ha ecognizes ha human beings ha e a
capaci y o ac , bu ha in doing so hey d aw on di e se logics o ende hei ac ions
meaning ul and consequen ial. Social ac o s may also be embedded in mul iple ins i u ional
logics, which ende s hei ac ions less p edic able and domina ed by a single logic (see e.g.,
G eenwood & Suddaby, 2006; Ba ilana & Do ado, 2010; Ezzamel, Robson & S aple on, 2012;
Lande , Koene & Linssen, 2013; Besha o & Smi h, 2014; Ca lsson-Wall, K aus & Messne ,
2016). Howe e , mos esea ch on ins i u ional logics has mainly a ended o he s uc u al
dimensions o ins i u ional change and has emphasized i s e olu iona y na u e as i un olds
o e longe pe iods o ime (Lounsbu y & Boxenbaum, 2013; Zilbe , 2013, 2017). Hence,
simila o ea lie ad ances in IT, he ins i u ional logics app oach ends o concei e o a leas
apid and adical change as a a he excep ional s a e ha is always condi ioned by ex an
ins i u ions. Mo eo e , shi s be ween ins i u ional logics ha e been seen as a ela i ely o de ly
p ocess, which has led o c i icisms ha much wo k in he a ea has e ec i ely subsc ibed o an
on ology which unde plays he complexi y in ol ed in he ep oduc ion and ans o ma ion o
ins i u ions (Qua one, 2015).
The mo e ac o -cen ic s and o IT comp ises se e al app oaches, p ima ily e ol ing unde
he ub ics o ins i u ional en ep eneu ship (DiMaggio, 1988), Scandina ian ins i u ionalism
(Cza niawska & Se on, 1996) and, mos ecen ly, ins i u ional wo k (Law ence & Suddaby,
2006). These app oaches ha e g own ou o he c i icisms o ea ly ad ances in IT o neglec ing
he social p ocesses h ough which human agen s ep oduce and ans o m ins i u ions, bu a y
in he emphasis placed on agency as an ins i u ionally embedded phenomenon. Resea ch on
ins i u ional en ep eneu ship ini ially eme ged as an a emp o e-ins a e a ocus on in e es -
d i en ac o s in IT, bu has been c i icized o subsc ibing o an unde -socialized, on ological
concep ion o ac o s as indi idual change agen s and he eby downplaying he p oblem o
collec i e agency and je isoning no ions o ins i u ional embeddedness (see Ha dy & Magui e,
2008; Ba ilana e al., 2009). G ea e conce ns wi h collec i e agency can be ound in
Scandina ian ins i u ionalism, which o eshadowed a shi in on ology owa ds a posi ion
which ecognizes he possibili ies o ins i u ional change as an ongoing and inde e mina e
p ocess which g adually shapes sha ed ideas and iden i ies in o ganiza ional ields
(Cza niawska & Se on, 1996; Cza niawska, 2008). In in oducing his pe spec i e,
Cza niawska & Se on (1996, p. 8) p opaga ed a iew o change as “a esul o a blend o
in en ions, andom e en s and ins i u ional no ms”, whils ecognizing ha any change in
o ganiza ional p ac ices is ine i ably condi ioned by ex an ins i u ions and ha i he e o e co-
exis s wi h a deg ee o s abili y. Howe e , mos empi ical wo k wi hin his a ian o IT has
ended o o eg ound he in ica e social p ocesses h ough which agen s con inuously
ans o m ins i u ions and has no eally ad anced he discussion o he on ological s a us o
ac o s as ins i u ionally embedded agen s (see Cza niawska, 2008; Sahlin & Wedlin, 2008;
Boxenbaum & S aandgaa d Pede sen, 2009). Mo e conce ed e o s o add ess he la e opic
a e disce nible in esea ch on ins i u ional wo k, which has sough o de elop an on ological
concep ion o indi idual and collec i e agency as in en ional and in e es -d i en wi hou
e e ing o an unde -socialized iew o ac o s (Law ence & Suddaby, 2006; Ba ilana &
D’Aunno, 2009). Empi ical wo k wi hin his s eam o esea ch has made impo an
con ibu ions o ou unde s anding o embedded agency (see e.g., Zie sma & Law ence, 2010;
Malsch & Gend on, 2013; Chiwami , Modell & Yang, 2014; Richa dson & Kil oyle, 2016)

6
and he o igina o s o he ins i u ional wo k app oach ha e epea edly emphasized he need o
concei e o agency as an ins i u ionally embedded phenomenon (see e.g., Law ence, Suddaby
& Leca, 2011; Hampel, Law ence & T acey, 2015).
Epis emological commi men s
By way o o e - iding epis emological posi ion, IT is based on a commi men o bo h induc i e
and deduc i e heo izing (Schneibe g & Clemens, 2006; G eenwood e al., 2008). The need
o induc i e esea ch was made plain in DiMaggio and Powell’s (1983) o iginal depic ion o
o ganiza ional ields as hey emphasized ha he exac na u e o such ields canno be
de e mined a p io i, bu needs o be disco e ed h ough empi ical inqui ies. Howe e ,
ins i u ional heo is s equen ly combine induc i ely gene a ed insigh s wi h a hypo he ico-
deduc i e mode o heo y de elopmen (Schneibe g & Clemens, 2006) and ha e con inuously
sough o ake s ock o eme ging esea ch indings in he o m o cohe en concep ual syn heses
(see e.g., Law ence & Suddaby, 2006; Sco , 2001; Tho n on e al., 2012). The deduc i e
elemen is also mani es in he ex ensi e bo owing and inco po a ion o concep s and ideas
om o he social heo ies as a means o illing concep ual gaps and o e coming limi a ions in
ins i u ional amewo ks. This bo owing o elemen s om o he heo ies has been an impo an
sou ce o inno a ion and has os e ed an e e -expanding heo e ical edi ice gea ed owa ds
explaining ins i u ional pe sis ence and change ac oss a ange o o ganiza ional phenomena
and empi ical se ings (McKinley, Mone & Moon, 1999; G eenwood e al., 2008; Boxenbaum
& Rouleau, 2011).
These e o s o con inuously e ine and ex end IT ha e led se e al commen a o s o a gue ha
i is i mly wedded o an essen ially ealis epis emology (Bow ing, 2000; Clegg, 2010), which
is unde pinned by a s ong no mal science adi ion gea ed owa ds con inuous heo e ical
puzzle-sol ing (Coope e al., 2008). These no mal science inclina ions ha e a guably led
ins i u ional heo is s o a ou heo e ical p ecision and elegance o e a emp s o explo e he
mo e mundane, ongoing and o en inde e mina e p ocesses h ough which ins i u ions a e being
( e-)p oduced on an e e yday basis. This endency was ein o ced by he dominance o
quan i a i e esea ch me hods in he ea ly de elopmen o IT (Schneibe g & Clemens, 2006)
and con ibu ed o he ei ica ion o ins i u ions as highly objec i ied and immu able en i ies
(Bow ing, 2000). Howe e , ecen c i iques sugges ha much o he ealis , no mal science
adi ion in IT has pe sis ed despi e o e wo decades o esea ch seeking o p oduce mo e
balanced accoun s o how ins i u ions in e ac wi h human agency and he inc easing
p opensi y o concei e o o ganiza ional ields as less s able and homogeneous en i ies (Zald
& Lounsbu y, 2010; Zilbe , 2013; Modell, 2015a; Willmo , 2015). Resea che s wo king
wi hin especially he Scandina ian ins i u ionalism and ins i u ional wo k adi ions ha e
del ed in o he complex social dynamics which ein o ce he inde e mina e na u e o
ins i u ionaliza ion. Ye , doub s emain as o whe he his has been accompanied by a shi
owa ds a widely accep ed epis emological posi ion which elaxes he impulse o con inuously
e ine and ex end IT. Tendencies in his di ec ion a e disce nible in Scandina ian
ins i u ionalism, whe e esea che s ha e concen a ed on p oducing ich, e hnog aphic
accoun s o how ins i u ional p ocesses un old a he han ad ancing elabo a e heo e ical
syn heses (Cza niawska, 2008; Sahlin and Wedlin, 2008). Howe e , his has a guably led his
s eam o esea ch o ha e ela i ely limi ed impac on o he s ands o IT (G eenwood e al.,
2008; Boxenbaum & S aandgaa d Pede sen, 2009), mos o which con inue o display a
p onounced no mal science impulse. This impulse is pa icula ly s ong in esea ch on
ins i u ional logics and ins i u ional wo k, which ha e bo h e ol ed in o expansi e esea ch
p og ammes held oge he by an epis emological commi men o con inuously e ining and
ex ending IT (see Law ence, Leca & Zilbe , 2013; Lounsbu y & Boxenbaum, 2013; Hampel
7
e al., 2015). This has a guably led o a pauci y o deepe , eal- ime analyses o he messy,
e e yday p ac ices in ol ed in ep oducing and ans o ming ins i u ions (Zilbe , 2013, 2017).
The p opensi y o much IT esea ch o emphasize he o de ly and p edic able na u e o
ins i u ionaliza ion has also been ein o ced by he ela i ely ixed and clea ly dema ca ed
assump ions abou key uni s o analysis, which ollow om i s poli ically conse a i e
epis emology (Coope e al., 2008; Clegg, 2010; Zald & Lounsbu y, 2010). The
epis emological bias agains mo e open-ended analyses is especially no able wi h espec o he
na ow concep ion o which ac o s ma e in o ganiza ional ields. Whils ypically ollowing
DiMaggio and Powell’s (1983) ad ice o induc i ely map ou he ela ions be ween a ious
ac o s in pa icula ields and ejec ing no ions o educ ionism (Schneibe g & Clemens, 2006;
Woo en & Ho man, 2008), ins i u ional heo is s ha e gene ally concen a ed hei inqui ies
o a ela i ely limi ed ange o powe ul ac o s a ou ed by ex an ins i u ions o exe ing a
dominan in luence on ins i u ional change (Law ence & Suddaby, 2006; Modell, 2015a). This
endency o ocus on a small numbe o eli e ac o s can be aced o DiMaggio and Powell’s
(1983) emphasis on he S a e and he p o essions as he main o igina o s o ins i u ional
pe sis ence and change in con empo a y Wes e n socie y. E en hough subsequen ad ances,
such as F iedland and Al o d’s (1991) in oduc ion o he ins i u ional logics pe spec i e,
a emp ed o loca e he o igins o ins i u ions wi hin wide sphe es o socie y, his ela i ely
na ow concep ion o ele an ac o s has con inued o pe mea e ins i u ional analyses. This has
os e ed a a he unques ioning epis emological app oach o he powe exe cised by social
eli es (Clegg, 2010; Zald & Lounsbu y 2010; Muni , 2015) and has led o epea ed calls o
widening he scope o ins i u ional analyses o include a b oade ange o cons i uencies and
examine how hei in e es s a e being p omo ed and hampe ed. Such calls ha e especially
eme ged in he li e a u e on ins i u ional wo k (e.g., Law ence & Suddaby, 2006; Law ence e
al., 2011), bu empi ical esea ch on he opic is s ill in i s in ancy (see Law ence e al., 2013;
Hampel e al., 2015) and has a guably been hampe ed by he no mal science impulse o
ins i u ional heo is s o con inuously ad ance heo y de elopmen a he han e lec ing mo e
deeply on hei epis emological p io s (Modell, 2015a; Willmo , 2015).
Ac o -ne wo k heo y
On ological ounda ions
In con as o IT, he on ological p emises o ANT
5
canno be as eadily elabo a ed wi h
e e ence o adi ional ca ego iza ions. Ins ead, ANT’s on ology seems bes cha ac e ized as a
mix o ela ionis , ealis and cons uc i is endencies, pi o ing on he key concep s o ac o ,
ne wo k and ansla ion. Whils he IT concep ion o an ac o is gene ally limi ed o human
beings (see Jones e al., 2013), in ANT an ac o is unde s ood as a mo e he e ogeneous and
con ingen en i y, “any hing ha has an e ec on ano he hing” (La ou , 2005, p. 71). In o he
wo ds, non-human en i ies a e conside ed as on ologically eal and as capable o exe cising
agency as human beings. This posi ion does no lead o a naï e an h opomo phic concep ion
o ma e ial a e ac s being endowed wi h simila capabili ies as human beings o o non-
humans’ ac ions being d i en by in en ions, bu simply ha he o m o agency does no in an
on ological sense ma e as long as i makes a di e ence in he su ounding wo ld (La ou ,
5
Ou e iew o he on ological and epis emological ea u es o ANT elies p edominan ly on he wo ks o La ou ,
Callon and Law as ANT is gene ally ecognized as ha ing eme ged especially om hei e o s (Blok & Jensen,
2011; Ha man, 2009). Fu he mo e, we ocus pa icula ly on he ideas o La ou as his wo ks o m he main sou ce
o inspi a ion o he as majo i y o ANT-inspi ed accoun ing esea ch (Jus esen & Mou i sen, 2011; Lukka &
Vinna i, 2014), including mos o he s udies unde e iew in his pape .
8
1987). The e o e, ANT s udies ha e always b immed wi h non-human ac o s such as scallops
(Callon, 1986a), labo a o y pa aphe nalia (La ou & Woolga , 1979), mic obes (La ou , 1988),
law sys ems (La ou , 2009) and ecological c ises (La ou , 2004). Simila ly, ANT-in o med
accoun ing esea ch has h own in o elie he agency o a ious accoun ing echnologies,
illus a ing o ins ance he eme gence o , and e ec s gene a ed by, calcula ions, ables, epo s,
pe o mance measu es and in o ma ion sys ems (e.g. P es on, Coope & Coombs, 1992;
Robson, 1992; B ie s & Chua, 2001; Mou i sen, Hansen & Hansen, 2009; Damb in & Robson,
2011; Qu & Coope , 2011; Vinna i & Skae baek, 2014; o name bu a ew).
The second undamen al concep o ANT, ha o he ac o -ne wo k, is de ined as
“simul aneously an ac o whose ac i i y is ne wo king he e ogeneous elemen s and a ne wo k
ha is able o ede ine and ans o m wha i is made o ” (Callon, 1987, p. 93). In o he wo ds,
an ac o eme ges om ela ional in e ac ions and i s cha ac e is ics a e ( e)de ined each ime i
is in ol ed in he dynamics o a ne wo k: “ he compe encies o he ac o will be in e ed a e
a p ocess o a ibu ion” (La ou , 1996a, p. 237; see also Callon & Muniesa, 2005, p. 1234).
Such ela ionis ideas a e mixed wi h ealis on ological belie s in he sense ha he ac o s and
ela ions which make up he wo ld a e conside ed o be eal a each momen . This iew o
objec s jus being he e, bo h independen o ou pe cep ion and in an a- heo e ical sense,
sugges s ha ANT’s on ological posi ion could e en be desc ibed as naï e ealism, a
cha ac e iza ion ha La ou (2005, p. 156) eadily accep s. Howe e , ANT de ia es om
classical ealism by a guing ha ac o s canno be di ided in o pe manen essen ial ea u es and
acciden al supe icial p ope ies; an ac o is simply he sum o all he p ope ies ha i has a a
pa icula momen (Ha man, 2009)
6
. Such an emphasis on ac ion and connec edness di e s
om he clea ly delinea ed ac o s and s able s uc u es p e alen in much IT esea ch. Ano he
di e ence o gene ic ealis iews is ha al hough ANT iews ex e nal eali y as being
independen o human ac ions and pe cep ions, i does no conside ha eali y o be
independen o he scien i ic de ices and me hods used o examine i (Law, 2004, pp. 31-32).
Wi h ce ain ools and da a collec ion p ocedu es, a ce ain eali y becomes cons uc ed,
implying ha se e al eali ies a e in ac possible
7
. Fo ins ance accoun ing ools, o mulas and
p og ammes oge he wi h hose who design and implemen hem o m socio- echnical
a angemen s ha pe o m he eali y ha hey measu e (Callon, 1998a, 2007). Such a iew
also displays a lack o dep h on ology, ha is, he belie ha eali y consis s o dis inc bu
embedded domains (Elde -Vass, 2008; O’Mahoney, O’Mahoney & Al-Amoudi, 2017), placing
ANT in s a k con as wi h he on ological hie a chy ha is a leas implici in IT.
In ANT lexicon, he p ocess o associa ions h ough which an en i y eme ges and acqui es i s
cha ac e is ics is known as ansla ion o cons uc ion, depending on which aspec o he
p ocess is emphasized. The no ion o ansla ion s esses he idea ha a ac o an inno a ion
does no eme ge as a esul o a linea de elopmen pa h bu is ans o med and modi ied along
i s unp edic able ajec o y. Impo an ly, i is no only he eme ging en i y ha is modi ied
du ing such a p ocess bu he p ope ies and in e es s o he ac o s connec ed o i a e
( e)de ined as well. Mo eo e , ansla ion always in ol es ials o s eng h in which he
pe sis ence o he objec s- o-be a e es ed by o he ac o s (Callon, 1986a); ha which is able o
esis such ials becomes “ eal” (La ou , 1987, p. 93). When in u n his p ocess is e e ed o
6
As is cha ac e is ic o ela ionism, hese p ope ies a e dynamic as hey a e con inuously ede ined in ne wo ked
in e ac ions, e en hough some o hem migh be mo e esis an o change han o he s.
7
Whe he hese di e se eali ies a e unde s ood as “plu al” o “mul iple” a ies be ween mo e adi ional ANT
accoun s and hose o he so called “ANT and A e ” school (e.g. Law & Hassa d, 1999; Law, 2004; Law &
Single on, 2005; Mol, 2002). Whils he o me ocus on examining how one e sion o eali y came in o being,
acknowledging ha al e na i e eali ies we e possible bu ne e enac ed, he la e a gue ha mul iple e sions o
eali y a e simul aneously pe o med h ough a ious p ac ices (e.g. Mol, 1999).
9
as “cons uc ion”, he aim is o highligh ha he eme gence o eal objec s is a cos ly, labo ious
achie emen ha equi es he collabo a ion o bo h humans and non-humans. Thus, ANT’s
“cons uc i ism” akes in o accoun a wide a ie y o ac o s han he human-cen ed “social
cons uc i ism” p e ailing in IT (La ou , 2005).
A dis inc i e ea u e o ANT’s on ology is i s ejec ion o pe manen and p e-exis ing duali ies,
whe he hese ela e o agency and s uc u e, na u e and socie y, o language and he wo ld.
Such duali ies a e conside ed o eme ge and become ( empo a ily) s abilized only a he end o
ansla ion p ocesses, a e a conside able amoun o wo k has been unde aken o c ea e
bounda ies be ween he di e en elemen s. Thus, in ma ked con as o IT, in ANT he no ion
o s uc u e as ypically de ined has no ole, excep as an objec o decons uc ion
8
, and hence
agency is no conside ed embedded ( o a p o ound c i ique o con en ional no ions o social
embeddedness see e.g. Callon, 1998a). ANT’s on ology is la , meaning, o ins ance, ha
s uc u es and o he mac o-ac o s a e no aken a p io i as some hing la ge han mic o-ac o s
such as indi iduals (La ou , 1996a). Ins ead o o ming a p e-exis ing, immu able con ex o
dynamic agency, s uc u es a e iewed analogously o any o he ac o s: hey a e momen a ily
s abilised agg ega es o local nego ia ions, con o e sies and o he in e ac ions in ol ing
humans and non-humans (Callon & La ou , 1981). Howe e , iewing he wo ld as
on ologically la does no cons i u e a denial o he exis ence o s uc u es o powe
di e en ials as such, bu he e-p esen a ion o bo h in an an i-essen ialis , ela ionis ligh . To
pa aph ase, s uc u al elemen s may be assumed o p e-exis analysis bu hen ANT is no he
app op ia e analy ical ool (La ou , 2005); IT o some o he social heo y can hen be employed
ins ead.
Whils IT unde s ands he wo ld o comp ise bo h s able elemen s (s uc u es) and change
s emming om ac i e agency, ANT conside s change o be ubiqui ous and in e wined wi h
he on-going ans o ma ion o ac o -ne wo ks. S abili y is seen as a a e, empo a y and agile
s a e he pe sis ence o which equi es cons an main enance (Callon, 1998b). Thus, ANT
s udies ocus mainly on he dynamics o change, pa icula ly on apid ans o ma ion ha o en
in ol es he eme gence o hyb ids, en i ies in which social and echnical elemen s a e
inex icably in e wined. In ac , in he absence o change o o he dynamics, ANT does no
ha e much o say in compa ison o mo e adi ional social heo ies, such as IT, which can hen
be employed o yield subs an i e explana ions. Howe e , adi ional heo ies may a imes be
inadequa e o accoun o ins ances in which bounda ies a e uzzy and i is unclea how a
pa icula en i y has been cons uc ed: “New opics, ha ’s wha you need ANT o !” (La ou ,
2005, p. 142). In line wi h his exho a ion, ANT schola s ha e examined, o ins ance, he
ad en o mic obiology and accina ion (La ou , 1988), he elec ic ehicle (Callon, 1986b),
ca bon ma ke s (Callon, 2009), and digi al na iga ion (La ou , 2010). Howe e , his emphasis
on inno a ion has also made ANT suscep ible o c i ique o a lack o sensi i i y o he
his o ical con ingency o social embeddness and he possibili ies o mo e endu ing o ms o
s uc u al s abili y (see e.g., Fou cade, 2007; Elde -Vass, 2008; Yang & Modell, 2015;
O’Mahoney e al., 2017).
Epis emological commi men s
Whils ins i u ional heo is s aim a e ining o ex ending heo y and se ou o do his by bo h
induc ion and deduc ion, ANT may a he be cha ac e ized as being a highly empi ical
app oach wi h no signi ican in e es in any o he men ioned heo iza ion s a egies o
easoning me hods. This posi ion de i es om ANT’s ambiguous na u e as nei he a ull-blown
8
In o he wo ds, an ANT schola can e y well conduc a e ospec i e analysis o how a momen a ily
s abilized s uc u e has been cons uc ed.
16
as hey a e se di e en le els o o ganiza ions (Adol sson & Wiks öm, 2007), policy
p og ammes (Ezzamel e al., 2007) and phases o public policy de elopmen (Hyndman e al.,
2014). Compa ed o adi ional no ions o ins i u ional isomo phism, hese s udies p o ide a
iche unde s anding o how accoun ing p ac ices come o a y as hey di use wi hin pa icula
ins i u ional se ings.
Howe e , simila o he mo e gene al de elopmen o Scandina ian ins i u ionalism,
accoun ing schola s ollowing his s and o IT ha e adop ed a highly ac o -cen ic app oach
o ins i u ional change whils la gely igno ing he iew o agency as an ins i u ionally
embedded phenomenon. Simila o much ANT-inspi ed esea ch on accoun ing (c . Jus esen
& Mou i sen, 2011; Lukka & Vinna i, 2014), he main emphasis o his esea ch is on how
a ious ac o s seek o de ise new accoun ing discou ses and p ac ices. E en hough he gene al
in luence o ex an ins i u ional s uc u es on ansla ion p ocesses is ecognized (see especially
Modell, 2005; Hyndman e al., 2014), no se ious e o s a e made o heo ise he possibili ies
o embedded agency. P obing in o hese possibili ies equi es much mo e de ailed a en ion o
how speci ic s uc u al mechanisms condi ion indi idual ac ion epe oi es and how such
mechanisms cons ain and enable he ans o ma ion o indi idual agency in o collec i e ac ion
(Ba ilana & D’Aunno, 2009; Modell, 2015a). Mo eo e , e en hough accoun ing esea ch
based on Scandina ian ins i u ionalism has d awn ex ensi ely on ANT-inspi ed no ions o
change, mos o i has honou ed he epis emological commi men s o IT, including he
concomi an no mal science impulse o con inuously e ine and ex end he la e me hod heo y.
This lea es he epis emological di e ences be ween IT and ANT and he ensions a ising om
hei di e ging iews o heo y de elopmen la gely un ecognized.
The emaining s udies in he “igno e” ca ego y make mo e angen ial use o ANT concep s,
bu ha e also sough o econcile such concep s wi h a s onge sense o ins i u ional
embeddedness. This includes examina ions o how budge s can be unde s ood as as a medium
h ough which ins i u ional logics shape o ganiza ional ac ion (Ezzamel e al., 2012) and how
a ious ins i u ional cons ain s shape he ansla ion o accoun ing epo s (Ca on & Tu co e,
2009; O’Neill e al., 2015). O pa icula in e es in his ega d is Ezzamel e al.’s (2012) s udy
o budge a y e o ms in he ield o educa ion. Following he ins i u ional logics app oach,
Ezzamel e al., (2012) complemen ed hei analysis wi h ANT-inspi ed no ions o
pe o ma i i y (Callon, 1998a) o enhance ou unde s anding o how budge ing is made o ac
as a ma e ial p ac ice and how his gi es ise o an icipa ed and unin ended consequences.
Budge s we e seen as an impo an media o be ween compe ing logics, exe cising
conside able in luence on e ol ing o ganiza ional p ac ices. Whils his is consis en wi h
ANT’s concep ion o how non-human ac o s come o ma e , Ezzamel e al. (2012) we e
ca e ul no o de ach his phenomenon om he ins i u ional con ex in which budge s we e
embedded. Thei s udy can hus be seen as an a emp o inco po a e conce ns wi h ma e iali y
and he ole o non-human ac o s in o IT, wi hou abandoning a iew o agency as an
ins i u ionally embedded phenomenon. Howe e , no explici a en ion is being paid o he
on ological ensions eme ging om such a emp s o econcile IT- and ANT-inspi ed no ions
o agency. The epis emological posi ion adop ed by Ezzamel e al. (2012) also ollows ha o
IT a he han ANT and displays p onounced conce ns wi h ex ending heo y de elopmen wi h
espec o how accoun ing is in luenced by ins i u ional logics.
A simila lack o a en ion o on ological ensions and an e en mo e p onounced absence o an
epis emological posi ion esembling ha o ANT a e disce nible in Ca on and Tu co e (2009)
and O’Neill e al. (2015). In e ms o on ology, bo h s udies ecognize how ex an s uc u es
condi ion he p opensi y o embedded agency (in he o m o pa h dependencies and
ins i u ional cons ain s on ansla ions). They also ecognize he capaci y o ma e ial

17
accoun ing p ac ices o cons i u e non-human ac o s and place ela i ely balanced emphasis on
he possibili ies o change and s abili y. Howe e , no a en ion is paid o he on ological
ensions esul ing om he econcilia ion o such ANT-inspi ed no ions o agency wi h he idea
o embedded agency. Also, in examining how accoun ing epo s ake shape, bo h Ca on and
Tu co e (2009) and O’Neill e al. (2015) adop an epis emological posi ion elying hea ily on
deduc i e heo izing o acili a e he analysis o a chi al da a. This de ac s om deepe and
mo e open-ended analyses o how a b oade ange o human and non-human ac o s a e
implica ed in inde e mina e change p ocesses. In bo h s udies, his led o a iew o change as a
highly cons ained phenomenon wi h a de ini e end-poin , which is a odds wi h ANT, bu he
au ho s ail o e lec on how hei lack o deepe a en ion o he unde lying change dynamics
may ha e ein o ced his concep ion o ins i u ionaliza ion. This illus a es how he adop ion
o an epis emological posi ion, which is much close o ha o IT han ANT, la gely nega es
he ambi ion o en ich ins i u ional analyses wi h insigh s om ANT. Howe e , he
epis emological ensions unde pinning his impasse emain un ecognized.
Taken oge he , ou analysis o he s udies in he “igno e” ca ego y d aws a en ion o he
di icul ies o combining IT and ANT in a logically cohe en way. In pa icula , he ensions
inhe en in he econcilia ion o IT- and ANT-inspi ed concep ions o agency and change seems
o p esen an on ological dilemma, which leads esea che s o ei he emphasize one o he o he
o hese concep ions. On he one hand, esea ch d awing hea ily on ANT o en ich analyses o
ins i u ional change, such as he s udies in o med by Scandina ian ins i u ionalism, ends o
downplay no ions o embedded agency. This b ings ins i u ional analyses close o he
concep ions o agency p e ailing in ANT, bu os e s ela i ely ague no ions o how ex an
ins i u ional s uc u es condi ion he possibili ies o agency and change. On he o he hand,
s udies which y o p ese e a s onge sense o how ins i u ions in luence agency and change
ha e la gely igno ed he ensions be ween such a iew and ANT’s open-ended no ion o change
as an ongoing and inde e mina e p ocess. Mo eo e , mos o he s udies in he “igno e”
ca ego y ollow he epis emological posi ion associa ed wi h IT wi hou e lec ing on he
jus i iabili y o bo owing concep s om ANT in ligh o hei di e ging iews o heo y
de elopmen . This is symp oma ic o he a he unques ioning no mal science endencies,
agains which c i ics o IT ha e inc easingly cau ioned (Coope e al., 2008; Willmo , 2015;
Modell, 2015a). As explica ed below, such endencies a e also e iden in pape s en ailing a
g ea e deg ee o e lec ion on he challenges o combining IT and ANT.
Assimila e
The pape s alling in o he “assimila e” ca ego y include e iew a icles deba ing in e alia he
possibili ies o combining IT and ANT (Lounsbu y, 2008; Modell, 2009) as well as wo
empi ical s udies pu suing a simila line o inqui y (Hoppe & Majo , 2007; A ena, A naboldi
& Azzone, 2010). The pape s classi ied in o his g oup a e all domina ed by IT, bu en ail mo e
explici ecogni ion o especially he on ological di e ences is-a- is ANT han hose alling
in o he “igno e” ca ego y. Howe e , he ex en o which his ecogni ion is accompanied by
deepe e lec ions on he pa adigma ic ensions esul ing om he combina ion o he wo
me hod heo ies a ies somewha .
S a ing wi h Lounsbu y’s (2008) discussion o how accoun ing p ac ices may be seen as
embedded in di e en ins i u ional logics, we see e idence o an a emp o p opaga e he use
o ANT as a complemen a y means o s udying such p ac ices wi hou abandoning IT as a
dominan me hod heo y. Consis en wi h IT, Lounsbu y (2008) explici ly subsc ibed o a
18
social cons uc i is on ology ecognizing how eali y is s uc u ed by ex an and eme ging
logics and how his condi ions he possibili ies o human agency and he p opensi y o
ins i u ional change and s abili y. Lounsbu y (2008, p. 357) ecognized ha such an on ological
posi ion is “seemingly a odds” wi h he mo e inde e mina e iew o change in ANT, bu a gued
ha he la e iew is “no an i he ical” o con en ional concep ions o ins i u ional change and
ha i can enhance ou unde s anding o he in ica e p ocesses h ough which a ia ions in
accoun ing p ac ices eme ge. This posi ion was jus i ied by he a gumen ha “ he e is always
change occu ing” (Lounsbu y 2008, p. 357) e en wi hin wha may appea o be e y
cons aining ins i u ional condi ions. Howe e , i is no clea how Lounsbu y (2008) sees ANT-
inspi ed analyses o change being econciled wi h no ions o embedded agency, which a e
s ongly emphasized in esea ch on ins i u ional logics. o whe he his will lead o a highly
ac o -cen ic concep ion o change simila o ha obse ed in Scandina ian ins i u ionalism.
Lounsbu y (2008) did no deepen his e lec ions on his opic o he on ological ensions a ising
om a emp s o econcile IT- and ANT-inspi ed no ions o agency and change.
Simila endencies o downplay on ological ensions a e disce nible in Modell’s (2009)
ex ension o Lounsbu y’s (2008) a gumen o he opic o pe o mance measu emen and
managemen . Adop ing an on ological posi ion simila o ha o Lounsbu y (2008), Modell
(2009) mobilized ANT as a means o nu u ing g ea e a en ion o pe o mance measu emen
as a ma e ial p ac ice and a mo e inde e mina e iew o change. In doing so, he did no see he
social cons uc i is on ology o IT as an insu moun able ba ie o he inco po a ion o ANT-
inspi ed insigh s al hough he ecognized he dis inc on ological o igins o he wo me hod
heo ies. Also, nei he Lounsbu y (2008) no Modell (2009) dwelled much on he
epis emological implica ions o combining IT and ANT in indi idual pieces o esea ch. Thei
e lec ions on he pa adigma ic ensions associa ed wi h such esea ch we e hus ela i ely
limi ed.
Somewha deepe e lec ions on he pa adigma ic ensions associa ed wi h combining IT and
ANT can be ound in he empi ical s udies alling in o he “assimila e” ca ego y. Bo h Hoppe
and Majo (2007) and A ena e al. (2010) paid explici a en ion o he on ological di e ences
be ween IT and ANT, bu d ew a he di e en conclusions as o how he wo me hod heo ies
migh be combined o ad ance a mo e open-ended iew o ins i u ional change as an ongoing
and inde e mina e phenomenon. Hoppe and Majo (2007) ecognized he di e ging, o e -
iding iews o eali y in IT and ANT, whils p ima ily leaning owa ds he o me me hod
heo y by aking he isomo phic p essu es embedded in ins i u ional s uc u es as a s a ing
poin o hei analysis. This was complemen ed wi h an ANT-inspi ed analysis o how he
egula ion o cos ing, based on an ac i i y-based cos ing app oach, was ansla ed wi hin an
indi idual o ganiza ion and how his caused cos ing p ac ices o de ia e om ield-le el
p esc ip ions. This mode o analysis is simila o ha p esc ibed by Lounsbu y (2008) in ha
i p o ides a ine-g ained depic ion o how p ac ice a ia ions can eme ge wi hin b oade
ins i u ional cons ain s. Howe e , simila o he accoun ing s udies inspi ed by Scandina ian
ins i u ionalism, Hoppe and Majo (2007) paid li le a en ion o he no ion o embedded
agency and he eby igno ed on ological ques ions as o whe he IT- and ANT-inspi ed
concep ions o agency can be combined in a logically cohe en way.
G ea e e o s o e ain a sense o embedded agency can be ound in A ena e al.’s (2010)
s udy o he ins i u ionaliza ion o isk managemen p ac ices. In con as o Hoppe and Majo
(2007), A ena e al. (2010) explici ly e-in e p e ed he on ology associa ed wi h ANT o
accommoda e a concep ion o human agency, which is close o ha o IT, in examining how
a ia ions in en e p ise isk managemen eme ged ac oss di e en o ganiza ions. In doing so,
hey ad anced he no ion o o ganiza ional change as an “embedded p ocess o ansla ion”
19
(A ena e al., 2010, p. 672) which is being condi ioned by mul iple ins i u ional logics. Whils
his may seem like an inno a i e way o concep ualizing he no ion o ansla ion i is, s ic ly
speaking, incompa ible wi h he on ological ounda ions o ANT. Hence, e en hough he
on ological di e ences be ween IT and ANT a e ecognized, he au ho s do no eally add ess
he ensions eme ging om he a emp o econcile hei di e ging iews o agency and
change.
Ou analysis o he s udies in he “assimila e” ca ego y ein o ces he iew ha esea che s
who seek o econcile IT- and ANT-inspi ed no ions o agency and change ace a a he
in ac able on ological dilemma. Simila o he s udies in he “igno e” ca ego y, accoun ing
esea che s seem compelled o ei he downplay no ions o embedded agency o subsc ibe o an
IT-inspi ed concep ion o agency which is inconsis en wi h ha o ANT. Mo eo e , e en
hough ANT concep s a e bo owed o nu u e a mo e open-ended analysis o change,
esea che s mainly ollow he epis emological commi men s o IT in a a he unc i ical manne .
The epis emological di e ences and ensions be ween IT and ANT a e ei he igno ed
(Lounsbu y, 2008; Modell, 2009) o subjec o ela i ely limi ed e lec ions cen ed on how a
mo e open-ended iew o ins i u ionaliza ion can be nu u ed (Hoppe & Majo , 2007; A ena
e al., 2010). Hence, simila o he s udies in he “igno e” ca ego y, he no mal science
aspi a ions o IT emain unques ioned and he issue o whe he his is compa ible wi h he mo e
a- heo e ical epis emological posi ion o ANT is le unadd essed.
D ama ize
The inal ca ego y o pape s includes mo e ex ensi e e lec ions on he pa adigma ic ensions
associa ed wi h combining IT and ANT and places ela i ely balanced emphasis on bo h
me hod heo ies as a basis o empi ical analysis. Only wo pape s (How & Alawa age, 2012;
Rau iainen & Scapens, 2013) all in o his ca ego y. As explica ed below, hese pape s also
di e signi ican ly in e ms o how hey app oach he combina ion o IT and ANT and deal
wi h pa adigma ic ensions.
How and Alawa age (2012) se ou o highligh he on ological di e ences be ween IT and
ANT and adop ed an explici ly eclec ic app oach o explain how new accoun ing p ac ices,
eme ging om he implemen a ion o an en e p ise esou ce planning sys em, emained
decoupled om ope a ions. In doing so, hey applied IT and ANT sepa a ely o o e wo
complemen a y accoun s o his phenomenon and made no e o s o in eg a e concep s om
one me hod heo y in o he o he . Consis en wi h he di e ging on ological posi ions
unde pinning hese me hod heo ies, IT-inspi ed no ions o decoupling as a s able s a e
embedded in ex an ins i u ional s uc u es we e jux aposed o an ANT-inspi ed accoun o how
ongoing ansla ion p ocesses and nego ia ions be ween a ious ac o s con ibu ed o main ain
his s a e. The need o such an analy ical app oach was jus i ied by he insigh ha he wo
me hod heo ies ha bou e y di e en , and no easily econcilable, iews o how
o ganiza ional change and decoupling a e b ough abou . The on ological ensions be ween he
wo pe spec i es can hus be said o ha e been ecognized and hen a oided by he choice o
analy ical app oach. This was acili a ed by he adop ion o a la gely induc i e, open-ended
epis emological posi ion which allowed wo sepa a e analyses o eme ge om empi ical da a
and which, acco ding o he au ho s, “ex ends he heo isa ion o decoupling” (How &
Alawa age, 2012, p. 404). To some ex en , his sepa a e use o IT and ANT lessened he
epis emological ensions associa ed wi h combining he wo me hod heo ies. Howe e , he
au ho s did no o e any deepe e lec ions on his opic and i is hus di icul o disce n
20
whe he hey ecognize he mo e undamen al epis emological di e ence be ween he wo
me hod heo ies ela ed o hei di e ging iews o heo y de elopmen . Thei mobiliza ion o
ANT o ex end ou unde s anding o a concep p ima ily associa ed wi h IT, such as decoupling
(c . Meye & Rowan, 1977), sugges s ha his is no he case and ha hei a emp o heo ize
his phenomenon is pe haps ye ano he example o he no mal science aspi a ions
cha ac e izing he la e heo y.
In con as o How and Alawa age (2012), Rau iainen and Scapens (2013) in eg a ed insigh s
om IT and ANT in o a uni ied amewo k and o e ed wha we belie e o be he mos
exhaus i e e lec ions on he pa adigma ic ensions associa ed wi h such esea ch o da e.
Ci ing se e al o he ea lie a emp s o combine IT and ANT (Hoppe & Majo , 2007;
Lounsbu y, 2008; Modell, 2009), hey sough o ake s ock o hese ad ances whils ex ending
hei analysis o he implemen a ion o an en e p ise esou ce planning sys em. In doing so,
hey ad anced ela i ely ex ensi e e lec ions on he on ological di e ences be ween IT and
ANT. E en hough he on ological posi ion guiding hei analysis leaned owa ds an IT-
inspi ed concep ion o ansla ion p ocesses as cons ained by ex an ins i u ional s uc u es,
hey clea ly ecognized he po en ial ensions associa ed wi h such a posi ion and he p oblems
o econciling i wi h a mo e inde e mina e iew o change. They also o e ed an insigh ul
empi ical accoun o how ins i u ional cons ain s condi ioned he ansla ion o en e p ise
esou ce planning and con ibu ed o a pa h-dependen p ocess o change.
Howe e , simila o se e al o he o he s udies included in ou e iew (e.g., Modell, 2005;
Adol sson & Wiks öm, 2007; Ezzamel e al., 2007; Hoppe & Majo , 2007; Hyndman e al.,
2014), Rau iainen and Scapens (2013) did no conside he mo e in ac able on ological issue
o whe he gene al concep ions o ansla ion as an ins i u ionally cons ained phenomenon can
be econciled wi h a de ailed unde s anding o embedded agency in a logically cohe en way.
No can hei analysis be said o ha e b oken wi h he no mal science aspi a ions o IT.
Rau iainen and Scapens (2013) paid signi ican a en ion o how IT migh need o be modi ied
o econcile i wi h he epis emological p inciples o ANT and nu u e an open-ended app oach
o ins i u ionalisa ion. Mo eo e , hey o e ed ex ensi e epis emological e lec ions on when
close in eg a ion be ween IT and ANT may be mos use ul and when he wo me hod heo ies
should be used in isola ion om each o he . In hei e o s o place he wo me hod heo ies
on a ela i ely equal oo ing, hey also paid explici a en ion o how IT may e ine and ex end
ANT. Howe e , he mo e undamen al jus i iabili y o econciling he epis emologies o he
wo me hod heo ies in ligh o hei di e ging iews o heo y de elopmen was le
unadd essed. No e e ences we e made o La ou ’s (1988, 1996b, 2005) p e e ence o ANT
s udies o always p og ess on a s and-alone basis wi hou esea che s imposing heo e ical
amewo ks on hose o be s udied, al hough he au ho s ecognized he gene al “dange o
o e - heo izing e en s” (Rau iainen & Scapens, 2013, p. 121) when combining heo ies.
The discussion abo e is indica i e o how ce ain on ological and epis emological ensions a e
di icul o a oid e en whe e esea che s display a ela i ely high deg ee o e lexi i y
conce ning he challenges o combining IT and ANT. As demons a ed by How and Alawa age
(2012), he only way o alle ia e some o hese ensions migh be o apply he me hod heo ies
sepa a ely o p o ide wo dis inc , bu complemen a y, accoun s o pa icula subs an i e
phenomena. Howe e , e en in hei case i is di icul o id onesel o he imp ession ha he
unde lying mo i a ion o using he wo me hod heo ies in andem o igina es in he no mal
science inclina ions, which ha e long cha ac e ized IT bu which a li e al eading o La ou
(1988, 1996b, 2005) de ies. Unless such li e al eadings o seminal ANT ex s a e elaxed, i
would seem o be i ually impossible o econcile he epis emologies o IT and ANT in a
logically cohe en way. Ye , as we ha e seen abo e, some esea che s s ill igno e his ension
21
and discuss how IT may e ine and ex end ANT in a manne which is, s ic ly speaking,
inconsis en wi h he epis emological commi men s o he la e heo y (see Rau iainen &
Scapens, 2013).
Concluding Discussion
This pape has que ied in o he challenges acing accoun ing esea che s who combine me hod
heo ies oo ed in di e ging on ological and epis emological assump ions h ough an analysis
o s udies which combine insigh s om IT and ANT. The majo i y o his esea ch is domina ed
by IT, whils inco po a ing concep s and ideas om ANT o ill pe cei ed gaps in he o me
me hod heo y. As such, ou analysis is mainly conce ned wi h a i mly es ablished body o
esea ch, which bo ows selec i ely om a pa icula me hod heo y, a he han adical
a emp s o de elop “new” heo ies h ough ull- ledged heo y blending (c . Oswick e al.,
2011; Suddaby e al., 2011). Ne e heless, simila o he gene al de elopmen o IT o e he
pas decades, accoun ing schola s ha e made claims o he e ec ha such bo owing can make
impo an con ibu ions o ou unde s anding o how accoun ing p ac ices a e ins i u ionalized.
The p ima y, o e - iding a ionale o combining IT and ANT has been ha he la e me hod
heo y can en ich he analysis o ins i u ional change as an ongoing and inde e mina e p ocess
and enhance ou unde s anding o he human and non-human agency in ol ed in he e olu ion
o he e ogeneous accoun ing p ac ices. This has, in u n, enabled accoun ing esea che s o
del e in o he dynamic na u e o powe ela ionships (e.g., Modell, 2005), he pe o ma i e
capaci y o accoun ing as a ma e ial p ac ice (e.g., Ezzamel e al., 2012), and he changing
meanings o accoun ing p ac ices (e.g., Adol sson & Wiks öm, 2007; Ezzamel e al., 2007;
Hoppe & Majo , 2007; A ena e al., 2010; Rau iainen & Scapens, 2013; Hyndman e al.,
2014), whils p ese ing a mo e o less p onounced sense o how such phenomena a e
condi ioned by he ins i u ional en i onmen in which hey e ol e.
Whils ecognizing hese subs an i e con ibu ions, we ha e demons a ed how he
combina ion o IT and ANT cons i u es a ela i ely ex eme example o me hod heo ies based
on incompa ible on ological and epis emological assump ions and how his gene a es se e e
pa adigma ic ensions. The mos signi ican ensions ela e o hei di e ging concep ions o
he na u e o social s uc u es and agency and hei e y di e en epis emological iews o he
ole o heo y. In e ms o on ology, much o he esea ch unde e iew seems o be caugh on
he ho ns o a a he in ac able dilemma. In dealing wi h he on ological ensions a ising om
he combina ion o IT and ANT, esea che s ei he ocus on social ac o s whils downplaying
no ions o embedded agency, which play an inc easingly p ominen ole in IT, o seek o
main ain a s onge sense o ins i u ional embeddedness, which is a odds wi h he concep ion
o agency in ANT. This endency o esea ch o bi u bica e along ac o -cen ic and
s uc u alis lines in explaining he p ocess o ins i u ionaliza ion is eminiscen o Coope e
al.’s (2008, p. 692) obse a ion ha esea che s ha e long “ lip- lop[ped] be ween ‘s uc u e’
and ‘agency’” in hei e o s o con inuously e ine and ex end IT. Ou indings also ein o ce
eme ging conce ns abou he di icul ies in econciling he on ology o ANT wi h heo ies
g ounded in a mo e s uc u alis unde s anding o he wo ld (Elde -Vass, 2008; O’Mahoney e
al., 2017). Howe e , he on ological ensions which eme ge om he combina ion o IT and
ANT ha e been subjec o li le explici e lec ion on he pa o esea che s. In addi ion, wi h
he excep ion o Qua one and Hoppe (2001), none o he s udies in ou e iew has explici ly
ecognized he di icul ies in econciling he no mal science aspi a ions embedded in IT wi h
he a- heo e ical epis emological commi men s o ANT. This lack o a en ion o key
epis emological ensions is pe haps less o a conce n, albei no unp oblema ic, as long as

22
esea ch is domina ed by IT and mainly un olds wi hin a no mal science adi ion (c .
O’Mahoney e al., 2017). Howe e , as we see e idence o in he s udies in he “d ama ize”
ca ego y, i becomes mo e p oblema ic when IT and ANT a e placed on an equal oo ing and
whe e his compels esea che s o also conside how he la e me hod heo y may be e ined
and ex ended.
No wi hs anding he pa adigma ic ensions associa ed wi h he combina ion o IT and ANT,
we do no wish o p onounce a mo a o ium on he e o s o imbue ins i u ional analyses wi h
a s onge sense o how powe ela ionships e ol e, how no ions o ma e iali y and non-human
agency a e implica ed in ins i u ional p ocesses and how he he e ogeneous meanings a ibu ed
o accoun ing p ac ices eme ge. Howe e , ou analysis aises ques ions as o whe he con inued
app ochemen be ween IT and ANT is he mos app op ia e a enue o his end o whe he
al e na i e pa hs, which a e no plagued by he same on ological and epis emological ensions,
can be ound. Whils a comp ehensi e discussion o his opic is beyond he scope o he p esen
pape , we wan o sugges a ew possible lines o inqui y o u u e esea ch. Two po en ially
p omising s ands o esea ch, which may be mo e amenable o econcilia ion wi h IT and i s
inc easing emphasis on embedded agency, can be ound in he li e a u es on socioma e iali y
(O likowski & Sco , 2008) and he pe o ma i i y o ma ke de ices (MacKenzie, 2006).
Whils bo h pe spec i es owe a conside able in ellec ual deb o ANT, hey a e s ill only loosely
connec ed o i s on ological ounda ions and a e a guably be e sui ed o examining he
e ol ing agen ial powe s o accoun ing as an ins i u ionally embedded phenomenon.
12
In
addi ion o hese bodies o esea ch, i may be wo h e isi ing accoun ing esea ch in o med
by he go e nmen ali y pe spec i e which, acco ding o Mille and Powe (2013), has
conside able a ini y o IT whils also inco po a ing insigh s om ANT. In pa icula , his
esea ch has a long-s anding in e es in how accoun ing plays a cons i u i e ole in
o ganiza ions and socie y, which is simila o he conce ns wi h pe o ma i i y, and i has
p o ided aluable insigh s in o how accoun ing p ac ices a e implica ed in a ecu si e in e play
wi h he ins i u ions ha bo h shape and a e being shaped by i s e olu ion.
Rega dless o which pa h is chosen o he u u e de elopmen o ins i u ional accoun ing
esea ch, we u ge esea che s o exe cise much g ea e e lexi i y wi h espec o he
on ological and epis emological p emises o such esea ch. Ou e iew e eals a spec um o
e lexi i y on he pa o esea che s. A small numbe o s udies, alling in o he “abandon” and
“d ama ize” ca ego ies, en ail ela i ely ex ensi e e lec ions on whe he he combina ion o
IT and ANT in a single s udy is jus i iable and how he pa adigma ic ensions associa ed wi h
doing so migh be add essed. Bu as we ha e also shown, he as majo i y o he pape s unde
e iew, alling in o he “igno e” and “assimila e” ca ego ies, include no o ela i ely limi ed
e lec ions on hese opics. We belie e esea che s could make he pa adigma ic implica ions
o hei wo k mo e explici by ca e ully spelling ou how he on ological ounda ions o he
me hod heo ies being combined o e lap o di e ge om each o he and how his eeds h ough
in o epis emological commi men s. This migh sensi ize esea che s o he ex en o which he
on ological and epis emological assump ions o a ious me hod heo ies can be econciled.
Such e lec ions do no necessa ily need o dis ac he p esen a ion o subs an i e esea ch
indings and heo e ical con ibu ions. Howe e , we also ecognize ha when me hod heo ies
a e as a apa in e ms on ological and epis emological assump ions, as is he case wi h IT
12
In he case o socioma e iali y, eseache s ha e long demons a ed a willingness o engage wi h ins i u ional
heo is s (O likowski & Ba ley, 2001) and ha e ecen ly e-cas he e ec s o socioma e ial p ac ices as a o m o
embedded agency (Leona di, 2013). Simila ly, MacKenzie’s (2006) wo k has been hailed as a use ul co ec i e
o ANT’s lack o sensi i i y o embeddedness, since he pays g ea e a en ion o how he pe o ma i i y o ma e ial
a e ac s is condi ioned by in e alia p e-exis ing social s uc u es (Fligs ein & Dau e , 2007; Fou cade, 2007).
23
and ANT, i may be p ac ically impossible o combine hem in a logically cohe en way. In
such ci cums ances, esea che s need o exe cise conside able es ain (Okhuysen & Bona di,
2011) and a oid he emp a ion o simply b acke on ological and epis emological assump ions
in hei e o s o in eg a e subs an i e heo e ical insigh s (Schul z & Ha ch, 1996). As
indica ed in he in oduc ion o his pape , such b acke ing o on ological and epis emological
assump ions seems o be ela i ely widesp ead in in e -disciplina y accoun ing esea ch and
has led o c i icisms ha accoun ing esea che s may be engaging in excessi e eclec icism (see
Modell, 2013, 2015a, 2015b). Simila o Schul z and Ha ch (1996), we see such c i icisms as a
cause o conce n, since i may unde mine he combina ion o me hod heo ies as a alid
schola ly endea ou .
A he same ime, we do no wish o oman icize he no ion o esea che e lexi i y as a ma e
o assuming an enligh ened epis emological posi ion, ee om any o he biases which hinde
he objec i a ion o knowledge (c . Lynch, 2000), and we ecognize ha he p opensi y o
esea che s o combine me hod heo ies and o e lec on such p ac ices is p o oundly shaped
by he epis emic communi ies in which hey a e embedded. Following Bou dieu (e.g.,
Bou dieu, 1988; Bou dieu & Wacquan , 1992), we accep ha he p oduc ion o scien i ic
knowledge is condi ioned by he ins i u ionalized belie s and p ac ices, o doxa, o such
communi ies. Howe e , we also ake his lead in a guing ha , as membe s o an epis emic
communi y, we ha e a collec i e esponsibili y o e lec on ecei ed esea ch p ac ices in an
a emp o objec i ize scien i ic knowledge o ma ion. Whils we a e conscious o he cha ge
agains Bou dieu o adop ing a hype -objec i is ic app oach o e lexi i y (Lynch, 2000), we
do no wan o ela i ize he p oduc ion o scien i ic knowledge o such an ex en ha
es ablished esea ch p ac ices go unques ioned only because hey ha e been accep ed as alid
by a pa icula epis emic communi y. Hence, we ecognize he need o no only e lec on he
pa adigma ic implica ions o combining pa icula me hod heo ies, bu also o engage in
deepe epis emic e lexi i y (Bou dieu & Wacquan , 1992) as o whe he he p ac ice o
combining such heo ies is jus i iable o no in a mo e gene al sense.
Inso a as he b oade , in e -disciplina y accoun ing esea ch p ojec is conce ned, we belie e
ha he e is a need o g ea e epis emic e lexi i y as o wha jus i ies he combina ion o
me hod heo ies and how his a ec s ou wo k as an epis emic communi y. I may be a gued
ha in e -disciplina y accoun ing esea ch is inc easingly s eeped in a doxa which a ou s a
no mal science adi ion aimed a cons an ex ension and e inemen o me hod heo ies o
ad ance subs an i e insigh s in o accoun ing as a social and o ganiza ional p ac ice (Vollme ,
2009; Richa dson, 2017). I is possible ha he ela i ely unques ioning app oach o he
combina ion o me hod heo ies, which we ha e documen ed, is a e lec ion o his b oade
end o a ou cumula i e heo y de elopmen a he han del ing in o he epis emic p emises
o such p ac ices. Such b acke ing o epis emic e lexi i y can be help ul as i may unleash
esea che s’ c ea i i y and gene a e impo an heo e ical ad ances. As no ed by Weick (1999),
an inabili y among esea che s o a leas empo a ily es ain hei e lexi i y can be pa alyzing
and can lead o a si ua ion whe e li le subs an i e p og ess in heo y de elopmen is being
made. Howe e , we a e wa y ha a lack o epis emic e lexi i y can also ein o ce endencies
owa ds excessi e eclec icism in an en i onmen whe e esea che s a e con inuously pushed o
ad ance inc emen al con ibu ions in a ne e -ending ques o ad ance me hod heo ies (Weick,
1999; Suddaby e al., 2011).
These conce ns ha e pa icula signi icance o esea che s employing he wo me hod heo ies
examined in he p esen pape . We see a pa icula isk o accoun ing esea ch in o med by IT
emaining suscep ible o c i icisms o excessi e eclec icism, unless he endency o bo ow
om o he me hod heo ies is combined wi h g ea e e lexi i y on he jus i iabili y o doing
24
so. O e he yea s, ins i u ional heo is s ha e a guably engaged in li le e lec ion on hei
esea ch p ac ices as an epis emic communi y (Coope e al., 2008; Cza niawska, 2008) and
he s ong no mal science adi ion pe mea ing in e -disciplina y accoun ing esea ch is also
unlikely o s imula e such e lec ions. Accoun ing schola s in o med by ANT ace a la gely
e e se dilemma. As we ha e demons a ed in his pape , a li e al eading o especially La ou
(1988, 1996b, 2005) pu s ANT a odds wi h any no mal science adi ion aimed a cumula i e
heo y de elopmen . Ye , accoun ing schola s using ANT a e p esumably subjec o he same
demands o ad ance inc emen al heo e ical con ibu ions as he es o he in e -disciplina y
accoun ing esea ch communi y. Fu he esea ch is equi ed in o how accoun ing esea che s
wi h a s ong commi men o ANT ha e esponded, and how hey could espond, o such
demands. I such analyses e eal aces o no mal science aspi a ions, simila o hose obse ed
in IT, hen i is legi ima e o aise conce ns abou how he doxa o he in e -disciplina y
accoun ing esea ch communi y compels esea che s o de ia e om key pa adigma ic
assump ions o he me hod heo ies employed. This would, in u n, ein o ce ou conce ns
abou he endencies owa ds eclec icism in his communi y.
Re e ences
Adol sson, P. & Wiks öm, E. (2007). A e quan i ica ion: quali y dialogue and pe o mance.
Financial Accoun abili y and Managemen , 23, 73-89.
Al esson, M., Ha dy, C. & Ha ley, B. (2008). Re lec ing on e lexi i y: e lexi e ex ual
p ac ices in o ganiza ion and managemen heo y. Jou nal o Managemen S udies, 45, 480-
501.
A ena, M., A naboldi, M. & Azzone, G. (2010). The o ganiza ional dynamics o En e p ise
Risk Managemen . Accoun ing, O ganiza ions and Socie y, 35, 659-675.
Ba ilana, J. & D’Aunno, T. (2009). Ins i u ional wo k and he pa adox o embedded agency.
In T.B. Law ence, R. Suddaby, & B. Leca (Eds.), Ins i u ional wo k: ac o s and agency in
ins i u ional s udies o o ganiza ions (pp. 31–58). Camb idge: Camb idge Uni e si y P ess.
Ba ilana, J. & Do ado, S. (2010). Building sus ainable hyb id o ganiza ions: he case o
comme cial mic o inance o ganiza ions. Academy o Managemen Jou nal, 53(6), 1419-1440.
Ba ilana, J., Leca, B. & Boxenbaum, E. (2009). How ac o s change ins i u ions: owa ds a
heo y o ins i u ional en ep eneu ship. The Academy o Managemen Annals, 3, 65-107.
Bax e , J. & Chua, W.F. (2003). Al e na i e managemen accoun ing esea ch – whence and
whi he . Accoun ing, O ganiza ions and Socie y, 28, 97-126.
Bea ie, V. (2014). Accoun ing na a i es and he na a i e u n in accoun ing esea ch: issues,
heo y, me hodology, me hods and a esea ch amewo k. B i ish Accoun ing Re iew, 46, 111-
134.
Be ge , P.S. & Luckmann, T. (1967). The social cons uc ion o eali y. New Yo k: Doubleday.
Besha o , M.L. & Smi h, W.K. (2014). Mul iple ins i u ional logics in o ganiza ions:
explaining hei a ied na u e and implica ions. Academy o Managemen Re iew, 38, 364-381.
Blok, A. & Jensen, T.E. (2011). B uno La ou : hyb id hough s in a hyb id wo ld. London:
Rou ledge.
25
Bou dieu, P. (1988). Homo Academicus. Camb idge: Poli y P ess.
Bou dieu, P. & Wacquan , L.J.D. (1992). An in i a ion o e lexi e sociology. Camb idge:
Poli y P ess.
Bow ing, M.A. (2000). De/cons uc ing heo y: a look a he ins i u ional heo y ha posi i ism
buil . Jou nal o Managemen Inqui y, 9, 258-270.
Boxenbaum, E. & Rouleau, L. (2011). New knowledge p oduc s as b icolage: me apho s and
sc ip s in o ganiza ion heo y. Academy o Managemen Re iew, 36, 271-96
Boxenbaum, E. & S andgaa d Pede sen, J. (2009). Scandina ian ins i u ionalism – a case o
ins i u ional wo k. In T.B. Law ence, R. Suddaby, & B. Leca (Eds.), Ins i u ional wo k: ac o s
and agency in ins i u ional s udies o o ganiza ions (pp. 178-204). Camb idge: Camb idge
Uni e si y P ess.
Boyd, R., Gaspe , P. & T ou , J.D. (1991). (Eds.). The philosophy o science. Camb idge, MA:
MIT P ess.
B ie s, M. & Chua, W.F. (2001). The ole o ac o -ne wo ks and bounda y objec s in
managemen accoun ing change: a ield s udy o an implemen a ion o ac i i y-based cos ing.
Accoun ing, O ganiza ions and Socie y, 26(3), 237-269.
Callon, M. (1986a). Some elemen s o a sociology o ansla ion: Domes ica ion o he
scallops and he ishe men o S -B ieuc Bay. In J. Law (Ed.), Powe , ac ion and belie : a new
sociology o knowledge? (pp. 196-223). London, Sociological Re iew Monog aph: Rou ledge
and Kegan Paul.
Callon, M. (1986b). The sociology o an ac o -ne wo k: The case o he elec ic ehicle. In: M.
Callon, J. Law, & A. Rip (Eds.), Mapping he dynamics o science and echnology (pp. 19-34).
London: The Macmillan.
Callon, M. (1987). Socie y in he making: The s udy o echnology as a ool o sociological
analysis. In W.E. Bijke , T.P. Hughes, & T.J. Pinch (Eds.), The social cons uc ion o
echnological sys ems: new di ec ions in he sociology and his o y o echnology (pp. 83-103).
Camb idge, MA: MIT P ess.
Callon, M. (1998a). In oduc ion: The embeddedness o economic ma ke s in economics. In
M. Callon (Ed.), The laws o he ma ke (pp. 1-57). Ox o d: Blackwell Publishe s.
Callon, M. (1998b). An essay on aming and o e lowing: Economic ex e nali ies e isi ed by
sociology. In M. Callon (Ed.), The laws o he ma ke s (pp. 244-269). Ox o d: Blackwell
Publishe s.
Callon, M. (2007). Wha does i mean o say economics is pe o ma i e? In D. MacKenzie, F.
Muniesa, & L. Siu (Eds.), Do economis s make ma ke s? (pp. 311-357). P ince on, NJ:
P ince on Uni e si y P ess.
Callon, M. (2009). Ci ilizing ma ke s: Ca bon ading be ween in i o and in i o
expe imen s. Accoun ing, O ganiza ions and Socie y, 34(3-4), 535-548.
Callon, M. & La ou , B. (1981). Unsc ewing he big Le ia han; o How ac o s mac os uc u e
eali y and how sociologis s help hem o do so? In K. Kno -Ce ina, & A.V. Cicou el (Eds.),
32
Qua one, P. (2015). Go e ning social o de s, un olding a ionali y, and Jesui accoun ing
p ac ices: a p ocedu al app oach o ins i u ional logics. Adminis a i e Science Qua e ly, 60,
411-445.
Qua one, P. & Hoppe , T. (2001). Wha does o ganiza ional change mean? Specula ions on a
aken o g an ed ca ego y. Managemen Accoun ing Resea ch, 12, 403-435.
Rau iainen, A. & Scapens, R.W. (2013). Pa h dependencies, cons ained ans o ma ions and
dynamic agency: an accoun ing case s udy in o med by bo h ANT and NIS. Quali a i e
Resea ch in Accoun ing and Managemen , 10, 100-126.
Ra iola, E. & No bäck, M. (2013). B inging echnology and meaning in o ins i u ional wo k:
making news a an I alian business newspape . O ganiza ion S udies, 34, 1171-1194.
Ribei o, J. & Scapens, R.W. (2006). Ins i u ional heo ies in managemen accoun ing change:
con ibu ions, issues and pa hs o de elopmen . Quali a i e Resea ch in Accoun ing and
Managemen , 3, 94-111.
Richa dson, A.J. (2017). The disco e y o cumula i e knowledge: s a egies o designing and
communica ing quali a i e esea ch, Accoun ing, Audi ing and Accoun abili y Jou nal (in
p ess)
Richa dson, A. & Kil oyle, E. (2016). Accoun ing ins i u ions as uce: he eme gence o
accoun ing in he go e nance o ansna ional mail lows, Accoun ing, O ganiza ions and
Socie y, 55, 32-47.
Robson, K. (1992). Accoun ing numbe s as ‘insc ip ions’: ac ion a a dis ance and he
de elopmen o accoun ing. Accoun ing, O ganiza ions and Socie y, 17(7), 685-708.
Rosenbe g, A. (2016). Philosophy o social science (5 h ed). Boulde , CO:Wes iew P ess.
Sahlin, K. & Wedlin, L. (2008). Ci cula ing ideas: imi ia ion, ansla ion and edi ing. In R.
G eenwood, C. Oli e , K. Sahlin & R. Suddaby (Eds.), The Sage Handbook o O ganiza ional
Ins i u ionalism (pp. 218-242). Thousand Oaks, CA: Sage.
Schneibe g, M. & Clemens, E. (2006). The ypical ools o he job: esea ch s a egies in
ins i u ional analysis. Sociological Theo y, 24, 195-237.
Schul z, M. & Ha ch, M.J. (1996). Li ing wi h mul iple pa adigms: he case o pa adigm
in e play in o ganiza ional cul u e s udies. Academy o Managemen Re iew, 21, 529-557.
Sco , W.R. (2001). Ins i u ions and O ganiza ions (2nd ed). Thousand Oaks, CA: Sage
Publica ions.
Seo, M-G. & C eed, W.E.D. (2002). Ins i u ional con adic ions, p axis, and ins i u ional
change: a dialec ical pe spec i e. Academy o Managemen Re iew, 27, 222-247.
Suddaby, R., Ha dy, C. & Huy, Q.N. (2011) Whe e a e he new heo ies o o ganiza ion?
Academy o Managemen Re iew, 36, 236-246.
Thompson, M. (2011). On ological shi o on ological d i ? Reali y claims, epis emological
amewo ks, and heo y gene a ion in o ganiza ions s udies. Academy o Managemen Re iew,
36(4), 754-773.

33
Tho n on, P.H. & Ocasio, W. (2008). Ins i u ional logics. In R. G eenwood, C. Oli e , K.
Sahlin & R. Suddaby (Eds.), The Sage handbook o o ganiza ional ins i u ionalism (pp. 99-
129). Thousand Oaks, CA: Sage.
Tho n on, P.H., Ocasio, W. & Lounsbu y, M. (2012). The ins i u ional logics pe spec i e: a
new app oach o cul u e, s uc u e, and p ocess. Ox o d: Ox o d Uni e si y P ess.
Tolbe , P.S. & Zucke , L.G. (1996). The ins i u ionaliza ion o ins i u ional heo y. In S.R.
Clegg, C. Ha dy, & W.R. No d (Eds.), Handbook o o ganiza ion s udies (pp. 175-190).
London: Sage Publica ions.
Vinna i, E. & Skae baek, P. (2014). The unce ain ies o isk managemen : A ield s udy on
isk managemen in e nal audi p ac ices in a Finnish municipali y. Accoun ing, Audi ing &
Accoun abili y Jou nal, 27(3), 489 – 526.
Vollme , H. (2009). Managemen accoun ing as no mal science. Accoun ing, O ganiza ions
and Socie y, 34, 141-150.
Weick, K.E. (1999). Theo y cons uc ion as disciplined e lexi i y: adeo s in he 90s.
Academy o Managemen Re iew, 24, 797-806.
Willmo , H. (2015). Why ins i u ional heo y canno be c i ical. Jou nal o Managemen
Inqui y, 24, 105-111.
Woo en, M. & Ho man, A.J. (2008). O ganiza ional ields: pas , p esen and u u e. In R.
G eenwood, C. Oli e , K. Sahlin & R. Suddaby (Eds.), The Sage handbook o o ganiza ional
ins i u ionalism (pp. 130-148). Thousand Oaks, CA: Sage.
Yang, C. & Modell, S. (2015). Sha eholde o ien a ion and he aming o managemen con ol
p ac ices: a ield s udy in a Chinese s a e-owned en e p ise. Accoun ing, O ganiza ions and
Socie y, 45, 1-23.
Zald, M.N. & Lounsbu y, M. (2010). The wiza ds o Oz: owa ds an ins i u ional app oach o
eli es, expe ise and command pos s. O ganiza ion S udies, 31, 963-996.
Zie sma, C. & Law ence, T.B. (2010). Ins i u ional wo k in he ans o ma ion o an
o ganiza ional ield: he in e play o bounda y wo k and p ac ice wo k. Adminis a i e Science
Qua e ly, 55, 189-221.
Zilbe , T. (2013). Ins i u ional logics and ins i u ional wo k: should hey be ag eed? Resea ch
in he Sociology o O ganiza ions, 39A, 77-96.
Zilbe , T. (2017). How ins i u ional logics ma e : a bo om-up app oach. Resea ch in he
Sociology o O ganiza ions, 48A, 137-155.
34
Table 1. Compa ison o he key on ological and epis emological assump ions o ins i u ional
and ac o -ne wo k heo ies.
Ins i u ional heo y
Ac o -ne wo k heo y
On ological ounda ions
O e - iding concep ion o
eali y.
The eal is socially cons uc ed, bu
ealis dimension also ecognized as
social s uc u es become objec i ied o e
ime (dep h on ology).
Whils he eal eme ges in a
ela ional and cons uc ed manne ,
objec s a e ully eal a any gi en
momen as in naï e ealism. An
ac o does no exis ou side i s
ela ions o o he s. All ac o s a e
on he same le el ( la on ology).
Concep ion o social
s uc u es.
S uc u es exis as socially eal en i ies
which condi ion, bu a e on ologically
dis inc om, human agency.
S uc u es a e no immu able, p e-
exis ing en i ies ha condi ion
agency. They can eme ge and
become empo a ily s abilized
only a he end o a ansla ion
p ocess.
Concep ion o agency.
Human agency is ins i u ionally
embedded and condi ioned by
his o ically con ingen s uc u es.
Inc easing a en ion o collec i e agency.
Agency is no embedded. Human
and non-human en i ies (ac an s)
con inuously in e ac and bo h
ha e he capaci y o exe cise
agency.
Concep ion o change and
s abili y.
Change is a ela i ely excep ional and
episodic s a e and is always condi ioned
by ex an ins i u ions.
Change is ubiqui ous and
in e wined wi h he on-going
ans o ma ion o ac o -ne wo ks.
S abili y seen as a empo a y and
agile s a e.
Epis emological commi men s
O e - iding epis emological
posi ion.
Empi ical/induc i e, bu also en ailing
impo an deduc i e elemen .
Highly empi ical, pu suing unique
desc ip ions and being open o all
possibili ies. Induc ion and
deduc ion do no play any
signi ican ole.
Role o heo y
No mal science iew aimed a
con inuous heo e ical e inemen and
ex ensions.
Theo y has no ole in ANT i sel ,
making he analysis inde e mina e.
P io assump ions abou uni s
o analysis.
Rela i ely ixed and clea ly dema ca ed
a p io i assump ions abou which ac o s
ma e , al hough he exac na u e o
ins i u ional en i onmen s can only be
disco e ed h ough empi ical
in es iga ion.
Delibe a ely open-ended abou
which ac o s and ela ionships
ma e in he o ma ion o ac o -
ne wo ks.
Table 2. Summa y o he e iew o accoun ing s udies combining IT and ANT.
S udy,
jou nal and
dominan
me hod
heo y.
On ological assump ions
Epis emological commi men s
O e - iding
concep ion
o eali y.
Concep ion o
social
s uc u es.
Concep ion o
agency.
Concep ion o
change and
s abili y.
O e - iding
epis emological
posi ion.
Role o heo y
P io assump ions
abou uni s o
analysis.
Abandon
Qua one
and
Hoppe
(2001),
MAR,
ANT
dominan .
Rejec ing
mode nis
iews o
social
eali ies as
s uc u ed by
objec i ied
ins i u ions.
Ex an
s uc u es
downplayed.
No a en ion o
embedded
agency whils
non-human
ac o s
ecognized.
Con inuous
change as
“d i ”
emphasized.
Explici ly open-
ended o a oid
p econcei ed
no ions o
p ac ice.
Li le a p io i
heo izing. ANT
p oposed as
al e na i e o IT.
Open-ended,
eme gen iew o
which ac o s and
ela ionships ma e .
Mennicken
(2008),
AOS,
ANT
dominan .
No made
explici .
Ex an
s uc u es
d i ing
di usion and
isomo phism
downplayed
bu in luence
o ex an ideas
ecognized.
No a en ion o
embedded
agency whils
non-human
ac o s
ecognized.
Ongoing and
inde e mi-na e
na u e o change
emphasized.
Rela i ely open-
ended al hough
in luence o
ex an ideas
ecognized.
Li le a p io i
heo izing. ANT
no ion o ansla ion
p oposed as
al e na i e o IT
no ion o di usion.
Open-ended,
eme gen iew o
which ac o s and
ela ionships ma e .
Igno e
Gend on
and Bake
No made
explici .
No a en ion o
ex an
s uc u es, bu
No explici
a en ion o
embedded
Main emphasis
on change/
c ea ion o ocal
Rela i ely open-
ended analysis bu
some p ede ined
ANT dominan bu
ex ended wi h no ions
Open-ended,
eme gen iew o
36
(2005),
EAR
ANT
dominan
dependence o
ocal ne wo k
on suppo ing
ne wo ks
ecognized.
agency o non-
human ac o s.
ne wo k bu
some a en ion
o s abilisa ion
o ne wo k.
heo e ical
concep s used.
o imi a ion and
ins i u ional-isa ion.
which ac o s and
ela ionships ma e .
Modell
(2005),
AAAJ,
IT
dominan
No made
explici .
A en ion o
ex an
s uc u es in
he o m o
ins i u ional
cons ain s.
No explici
a en ion o
embedded
agency o non-
human ac o s.
Main emphasis
on change bu
also a en ion o
s abilising ole
o esis ance.
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
IT dominan bu
ex ended wi h no ion
o ansla ion o
unde s and
powe /poli ics.
Rela i ely open-
ended iew o which
ac o s ma e , bu
con ined o ocal
o ganiza ional ield.
Adol sson
and
Wiks om
(2007),
FAM,
IT
dominan
No made
explici .
A en ion o
ex an
s uc u es in
he o m o
exis ing
iden i ies.
No explici
a en ion o
embedded
agency o non-
human ac o s.
Ambi ion o
s udy bo h bu
main emphasis
on change
(excep o
s abilising
e ec s o
decoupling).
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
IT dominan and
ex ended wi h no ion
o ansla ion.
Key o ganiza ional
ac o s la gely
p ede ined.
Ezzamel e
al. (2007),
AAAJ,
IT
dominan
No made
explici .
Ex an
s uc u es
ecognized bu
main emphasis
on c ea ion o
new
ins i u ions.
No explici
a en ion o
embedded
agency o non-
human ac o s.
Change
emphasized.
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
IT dominan and
ex ended wi h no ion
o ansla ion o
unde s and p ocess o
change.
Key ield-le el ac o s
la gely p ede ined.
37
Ca on and
Tu co e
(2009),
AAAJ,
IT
dominan
No made
explici .
Ex an
s uc u es and
hei e ec s on
pa h
dependencies/
c ea ion
ecognized.
Some
concep ual
a en ion o
embedded
agency and
ma e iali y bu
no ecei ing
empi ical
a en ion.
Concep ual
aming
ecognizing
bo h bu
empi ical
analysis
unde lining
s abilising ole
o ins i u ions.
Mos heo e ical
concep s p e-
de ined bu
combina ion o
deduc i e and
induc i e
app oaches.
IT dominan . IT-
inspi ed no ion o
agency jux aposed o
ANT o explain pa h
dependence/
c ea ion.
Key ac o s p e-
de ined (ins i u ional
en ep eneu and
adop e s).
Ezzamel e
al. (2012),
AOS,
IT
dominan .
No made
explici .
In e play
be ween
ex an and
eme ging
s uc u es in
he o m o
ins i u ional
logics
ecognized.
Explici
a en ion o
agency as
embedded in
ins i u ional
logics and he
pe o ma i e
ole o budge s
as ma e ial
p ac ice.
Rela i ely
balanced
emphasis on
bo h.
Deduc i e
de i a ion o
heo e ical
concep s
combined wi h
mo e induc i e
analysis.
IT dominan bu
ex ended wi h ANT-
inspi ed no ions o
pe o ma i i y o
de elop a sense o
agency.
Key ield-le el and
o ganiza ional ac o s
p e-de ined
Hyndman
e al.
(2014),
No made
explici .
A en ion o
ex an
s uc u es in
he o m o
sedimen s.
No explici
a en ion o
embedded
agency o non-
human ac o s.
Main emphasis
on change in
discou se bu
also a en ion o
con inui y.
Deduc i e
de i a ion o
heo e ical
concep s and
ca ego ies o
IT dominan bu
ex ended wi h no ion
o ansla ion.
O ganiza ional ields
and le els o analysis
p ede ined.

38
CPA,
IT
dominan
discou se p e-
de ined.
O’Neill e
al. (2015),
AAAJ,
IT
dominan
No made
explici .
Ins i u ional
s uc u es and
isomo phism
condi ioning
scope o
ansla ions.
Some
concep ual
a en ion o
embedded
agency and
ma e iali y bu
li le empi ical
a en ion o
pe o ma i i y.
Concep ual
aming
ecognizing
bo h bu
empi ical
analysis
unde lining
s abilising ole
o ins i u ions.
Highly deduc i e
heo izing
en ailing a p io i
hypo hesis
de elopmen .
IT dominan and used
o ame ANT
concep s o
insc ip ion and
ansla ion.
Uni o analysis
p ede ined bu some
a emp o
p oblema ize ex an
classi ica ion o
ou comes.
Assimila e
Hoppe
and Majo
(2007)
EAR,
IT
dominan .
Di e ging
on ologies in
IT and ANT
ecognized.
A en ion o
ex an
s uc u es as a
sou ce o
isomo phism.
Li le explici
a en ion o
embedded
agency o non-
human ac o s.
Change
emphasized bu
s abilising ole
o esis ance
also ecognized.
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
IT dominan , bu
complemen ed wi h
ANT o explain
inde e mina e na u e
o change p ocess.
O ganiza ional ields
and le els o analysis
p ede ined bu open-
ended iew as o
which ac o s ma e .
Lounsbu y
(2008),
AOS,
Explici
subsc ip ion
o social
cons uc i -
is on ology.
Emphasis on
in e play
be ween ex an
and eme ging
ins i u ional
logics.
Explici
a en ion o
embedded
agency.
Rela i ely
balanced
emphasis on
bo h.
Analy ical
concep s clea ly
de ined bu also
emphasis on
induc i e
esea ch.
IT dominan bu
ex ended wi h ANT
no ions o ne wo ks
and pe o ma i i y.
La gely p ede ined.
39
IT
dominan
Li le a en ion
o ma e iali y.
Modell
(2009),
FAM,
IT
dominan
Explici
subsc ip ion
o social
cons uc i -
is on ology.
Emphasis on
in e play
be ween ex an
and eme ging
ins i u ional
logics.
Explici
a en ion o
embedded
agency and non-
human ac o s.
Rela i ely
balanced
emphasis on
bo h.
Analy ical
concep s clea ly
de ined bu also
emphasis on
induc i e
esea ch.
IT dominan bu
ex ended wi h ANT
concep s o
ma e iali y and
ongoing change
p ocesses.
La gely p ede ined.
A ena e
al. (2010),
AOS,
IT
dominan
Explici e-
in e p e a io
n o he
on ology o
ANT o
ecognize
p e-exis ing
s uc u es.
Explici
a en ion o
ins i u ional
embedded-
ness
A en ion o
“embedded
p ocesses o
ansla ion”.
Some a en ion
o ma e ial
p ac ices bu
li le a en ion o
hei
pe o ma i i y.
Rela i ely
balanced
emphasis on
bo h (bu
p ocess o
change la gely
b acke ed).
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
IT dominan , bu
ex ended wi h no ion
o ansla ion.
Uni s o analysis
p ede ined bu open-
ended iew as o
which ac o s ma e .
D ama ize
40
How and
Alawa age
(2012)
CPA,
balanced
emphasis
on IT and
ANT.
Sepa a ion
o IT and
ANT
signi ying
ecogni ion
o di e en
on ologies.
Recogni ion o
ex an
ins i u ional
s uc u es as
pa o IT
analysis.
No a en ion o
embedded
agency.
Ex ensi e
discussion o
ma e iali y/
pe o ma i i y
as pa o ANT
analysis.
A en ion o
change bu
s abili y
emphasized due
o obus ness o
decoupling.
La gely induc i e
analysis o
gene a e sepa a e
heo e ical
accoun s.
IT and ANT applied
sepa a ely o ad ance
al e na i e
explana ions o
decoupling.
Uni o analysis
p ede ined bu
ela i ely open-ended
as o which ac o s
and ela ionships
ma e .
Rau iainen
and
Scapens
(2013),
QRAM,
balanced
emphasis
on IT and
ANT.
Ex ensi e
discussion o
di e ences
in on ology.
Ex an
ins i u ional
s uc u es seen
as cons ain s
on ansla ion
p ocesses.
Embedded
agency implici
in no ion o
“cons ained
ans-
o ma ions”, bu
no explici ly
mobilized.
Some a en ion
o ma e iali y.
Rela i ely
balanced
emphasis on
bo h bu
empi ical
indings
emphasizing
s abili y (pa h
dependence)
Deduc i e
de i a ion o
heo e ical
concep s bu
la gely induc i e
analysis.
Equal emphasis on IT
and ANT and
ex ensi e discussion
o simila i ies and
di e ences.
Uni o analysis
p ede ined bu
ela i ely open-ended
as o which ac o s
and ela ionships
ma e .