PRINCIPIA 28(4): 517–540 (2024) doi: 10.5007/1808-1711.2024.e92531
Published by NEL — Epis emology and Logic Resea ch G oup,Fede al Uni e si yo San a Ca a ina(UFSC),B azil.
COMPLEX NETWORKS, STRUCTURAL EXPLANATIONS,AND
THE ROLE OF VALUES IN EXPERIMENTAL LINGUISTICS
JUAN BAUTISTA BENGOETXEA
Uni e sidad del País Vasco/Euskal He iko Unibe si a ea, SPAIN
[email p o ec ed]
h ps://o cid.o g/0000-0003-1158-1122
Abs ac . Fi s , a b ie s a e o he a on he ole o alues in science is p esen ed. The main
poin s o he c i ique o he ideal o a ee- alue science a e depic ed (Sec. 2), ollowed by
a amewo k o ou phases in he de elopmen o scien i ic ac i i y ha allows us o be e
assess he in luence o alues in science. The basic hesis o he p esen ex c i icizes he
concep ion close o Hea he Douglas’ hesis, acco ding o which alues, including cogni i e
alues, can only play an indi ec ole in science (Sec. 3). In o de o show ha , a ce ain
phases o science, he explana o y powe (a cogni i e alue) can play a di ec ole, we p opose
an examina ion o s uc u al explana ion, as dis inguished om mechanis ic explana ion, in
a case o expe imen al linguis ics ela ed o language acquisi ion (Sec. 4). The conclusion is
p esen ed in he las sec ion.
Keywo ds: complex ne wo ks •s uc u al explana ion •expe imen al linguis ics •cogni i e
alues
RECEIVED: 18/01/2023 REVISED: 08/12/2023 ACCEPTED: 22/05/2024
1. In oduc ion
The social ele ance o philosophy con inues o be a opic o deba e in ou discipline.
Acco ding o Ki che (2023, p.27), he philosophy o science and he epis emology o
science would be qui e well placed on his landscape, which un o una ely shows a
shi away om public in e es in philosophical p oduc s. Haack (2016) a gues ha
in e es could be e i ed p ecisely i we we e o place philosophy in con ex s o mo e
social ele ance. In his espec , she ag ees wi h he esul s p ojec ed by au ho s such
as Douglas (2009b, 2014), Lacey (2004, 2011), Ma ins Dos Reis and De Pa a Pilla
(2018), o men ion a ew cases o a philosophy di ec ly linked o issues whose in e es
goes beyond he bounda ies o he academy.
Ob aining and managing eliable knowledge is an inhe en human goal and i s
social ele ance is ac ually e y impo an . I we conside ha he highes o m o do-
ing his is he scien i ic one, i s p oduc s, gene ally speaking ( heo ies, hypo heses,
models, echnologies), will ha e o be eliable o be used in ou ask o unde s anding
© 2024 The au ho (s). Open access unde he e ms o he C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
518 Juan Bau is a Bengoe xea
phenomena, explaining and p edic ing ac s and hypo heses, in e ening and in lu-
encing esea ch domains wi h a minimum gua an ee, o con olling a ious a iables
in hese domains (expe imen a ion) (see Hacking 1983, p.149–52; Douglas 2009b,
p.93–4, 100–2, 112–3; Ca w igh e al. 2022). In he case we a e conce ned wi h
he e, scien i ic explana ion, i s a ge s (explananda) depend o a la ge ex en on all
hese p oduc s o science and also on he alues unde lying he scien i ic en e p ise
in si ua ions—all o almos all—in which unce ain y and isk a e pa and pa cel o
he wo k ha mus be done (Douglas 2014). In hese e y common ci cums ances,
esea che s, egula o s and science policy make s ha e o choose ce ain esea ch
pa hways om a wide ange o op ions. I is he e o e e y usual o wonde abou
he c i e ia, he in e es o he objec i e ha makes scien i ic esea ch ocus on one
objec o ano he , on one s a egy o ano he , wi h such social weigh .
Any esponse will undoub edly ake he o m o a alue judgemen o depend on
alue judgemen s.1The ideal o a alue- ee science had a majo impac on he philos-
ophy o science and he epis emology o science o decades (see Rezaee & Bika aan-
Behesh 2023). Howe e , ha ideal was pu in o c isis and a majo shi owa ds he
examina ion o non-epis emic and non-cogni i e alues was unde aken (see Lacey
1999, 2004; Douglas 2009b). This change led o he almos consensual endo semen
o he belie ha e hical and social alues ( hese would be he non-epis emic/non-
cogni i e ones) do in luence scien i ic ac i i y. Cu en issues as impo an o science
and socie y as egula ion and decision making, in addi ion o d awing on scien i ic
wo k based on da a and e idence, mus ake in o accoun he in luence o alues and
in e es s o he s akeholde s in ol ed.2
The shi away om he ideal o a alue- ee science has, howe e , led us o
adically opposed posi ions, o he ex en ha he ole o epis emic and cogni i e
alues as such has been la gely downplayed. I seems ha he shi owa ds he s udy
o e hical and social alues has been excessi ely skewed owa ds ce ain biases and
e en posi ions ha deny o g ea ly educe he ole o epis emic and cogni i e alues
(Ha ding 1986, Longino 1990, see Haack 1998). This could call in o ques ion he
e y na u e o he scien i ic en e p ise. F om a mode a ely p o-epis emic and p o-
cogni i e app oach (see Risjo d 2023), we conside i impo an o pa ially e u n
o a de ence o he posi i e ole o epis emic and cogni i e alues, especially in a
human science such as linguis ics (Sec. 4).
To ponde he plausibili y o he hesis ha e hical and social alues domina e in
he social and human sciences, we i s look a an epis emological p oposal, speci ic
o he philosophy o science, which p ojec s an image o science as cons i u ed in
ou phases (Lacey 1999, Ma ins Dos Reis & De Pa a Pilla 2018, Douglas 2009b).3
I is wi hin hese phases ha we can inqui e in o he ole o epis emic, cogni i e and
social alues, and in o which, i any, a e s onges in each phase. The in e es o his
lies in he ac ha we a e in e es ed in assessing scien i ic ac i i y in a social and
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 519
human science such as linguis ics, a ield in which, acco ding o he ad oca es o he
ole o social and e hical alues, hese acqui e e en g ea e impo ance, i possible.
Howe e , he e we wan o a gue ha i is p ecisely epis emic and cogni i e al-
ues, in pa icula he explana o y capaci y, ha play a c ucial ole in his science,
and ha hey dispel some o misunde s andings su ounding he mul iplici y o in-
e p e a ions and he weakness o he e idence used. Building on Douglas’s (2009a,
2009b) dissec ion o he (di ec and indi ec ) ole o alues in science, ou analysis o
he case o s uc u al explana ion in linguis ics seeks o nuance ha esponse and o
e-emphasize he di ec ole o some cogni i e alues in scien i ic ac i i y, especially
ha o explana o y powe (see Ca w igh e al. 2022).
2. A science wi h social alue
Despi e he a o emen ioned shi owa ds he s udy o non-epis emic and non-cog-
ni i e alues, i is s ill possible o a gue ha he e a e epis emic and cogni i e al-
ues ha in luence scien i ic ac i i y in a e y ele an way. Hea he Douglas (2009b,
p.92) s a s om an in e es ing ypology o alues. Acco ding o he , we can dis in-
guish be ween e hical alues and social alues on he one hand, and epis emic alues
and cogni i e alues, on he o he . The la e wo ypes di e in ha epis emic al-
ues as such a e no eally alues, bu c i e ia ha any scien i ic en e p ise mus mee .
These a e es ablished because science is alued as such. They he e o e ix he scope
o science om he ou side, and do no ac as alues wi hin science. Cogni i e al-
ues, on he o he hand, cons i u e some hing mo e p ecise han a loose g ouping o
accep able alues in science ( his is wha is o en equa ed wi h epis emic alues).
Cogni i e alues, he poin s ou , a e hose elemen s o scien i ic wo k ha helps us o
e lec on he e iden ial and in e en ial ea u es o heo ies and da a ha scien is s
hold (Douglas 2009b, p.93). They would be assis an s o he cogni i e wo k o sci-
en is s. In sho , i cogni i e alues a e conce ned wi h he ui ulness o esea ch,
epis emic alues a e conce ned wi h he s a ing o basic goal o esea ch, namely he
ue o , mo e usually, eliable knowledge. Fu he mo e, and impo an ly, cogni i e,
social and e hical alues should ne e claim o play a di ec ole in science. Tha is,
hey should no be equa ed wi h he e idence and da a used by scien is s.
We see, he e o e, ha a alue o high epis emic incidence is he eliabili y o
he p oduc s o science. Ca w igh e al. (2022, Pa I, 3), in he wake o an idea o
O o Neu a h, poin ou ha eliabili y is an umb ella concep o Ballung. The con-
cep s linked o cogni i e alues (scope, simplici y, explana o y capaci y, consis ency)
would be deployed alongside i .4Science, among o he hings, would seek o p opose
cohe en models, hypo heses o heo ies, o o o e adequa e o use ul heo ies. Bu
we also know ha heo ies and he scien i ic p ac ices associa ed wi h hem ha e
PRINCIPIA 28(4): 517–540 (2024)
520 Juan Bau is a Bengoe xea
deg ees o unce ain y and a e allible and p o isional (see Sh ade -F eche e 1993),
a allibili y ha scien i ic p ac ices and heo ies seek o educe. Wha does seem o
be accep ed is ha he e is no such hing as a alue- ee science (whe he o no ha
ideal is upheld), on he one hand, and ha , al hough i is alue-laden, science s ill
has i s own i uous ea u es ( eliabili y, u h) wi hou which science would cease o
be science and esea ch would cease o be esea ch. In addi ion o hose men ioned
by Douglas (see endno e 4), we ind p ecision, cla i y, jus i ica ion, objec i i y, u h,
empi ical adequacy (see an F aassen 1980; Lacey 1997, p.7), espec o e idence,
as well as he e o o minimize allibili y and unce ain y (Douglas 2009b, p.95).
E hical and social alues also in luence science. E hical alues a e conce ned wi h
wha is good o igh , and a e e y impo an in examining he consequences o sci-
en i ic e o s o he gene al public. Among o he hings, hey help us o weigh up
whe he po en ial bene i s a e ac ually po en ial dange s o be aken in o accoun
o no , whe he he e a e dange s o be assumed wi hou any p ice, e c. (Douglas
2009b, p.92). Social alues a ise om wha a pa icula socie y alues: jus ice, p i-
acy, eedom, social s abili y, inno a ion, e c. (Lacey 1999, p.28–9).
Taking hese di e en ypes o alues se iously has p omo ed he a o emen ioned
c i ique o he ideal o a alue- ee science (Kincaid; Dup é; Wylie 2007), which has
changed he o e all scope o he deba e. Bo h he sup emacy o he old alue- ee
ideal and he exclusi e in e es in epis emic alues (McMullin 1982; Laudan 1984)
o scien i ic esea ch ha e been called in o ques ion (Douglas 2009b, p.89). A he
p esen i is social, poli ical o e hical alues ha a e o conside able in e es in e-
cen axiological deba es (Koskinen & Rolin 2022; McMullin 1982). The e o s o
o e come he ideal o a science ee o alues has been made in o de o adop an
inc easingly plu alis ic scien i ic a i ude, bo h in na u al and social sciences (Kelle ;
Longino; Wa e s 2006). This new a i ude is based on he basic idea ha scien i ic
esea ch is a human ac i i y ha is goal-o ien ed and highly in luenced by non-
epis emic alues and no ms o esea ch se ings (con ex s) (González 2013, p.1505).
The ou mos p ominen lines o c i icism o he ideal o a alue- ee science can be
summa ized as ollows:
[1]Philosophe s sough o ‘unmask’ he so-called epis emic alues ha a e in ac
non-epis emic ones. Longino (1996) has a gued ha so-called ‘epis emic’ alues a e
no in ac exclusi ely epis emic, o hei use assumes poli ical and social alues in
con ex s o scien i ic judgemen . The e a e alues ha unde lie scien i ic judgemen
and hey a e no always, o e en usually, poli ically neu al. Longino p oposes o
jux apose hese alues wi h alues (including many o a eminis na u e) such as
no el y, on ological he e ogenei y, mu uali y o in e ac ion, applicabili y o human
needs and di usion o powe . Failu e o ecognise his axiological ichness and a ie y
can lead o bias and ad e se esea ch ou comes (Wilhol 2009, p.96).
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 521
[2]Richa d Rudne ’s (1953) now classic c i ique (see Resnik & Ellio 2023, Lacey
1999, p.71–4) ques ioned he idea ha any decision abou a scien i ic heo y o hy-
po hesis can be alue- ee (Douglas 2009b, p.51-2). Acco ding o him, any accep-
ance o ejec ion o a hypo hesis is se agains a backg ound o knowledge and p i-
o i ies (Rudne 1953, p.2), and a scien i ic hypo hesis is ne e con i med beyond
easonable doub : he e is always he possibili y o a decision being w ong (Resnik
& Ellio 2023, p.7). The decision o accep o ejec a hypo hesis implies a alue
judgemen (a leas implici ly); one has o judge which o he consequences o an
e oneous decision is mo e accep able.5Hence, non-epis emic (mo al, social) alues
a e a necessa y pa o he scien is ’s co e ac i i y o accep ing and ejec ing hypo he-
ses.
[3]Mo eo e , using a s a egy simila o Longino’s (1996), Pu nam (2002, p.140–
2) ejec s he possibili y o a alue- ee science by means o a seman ic a gumen ha
a acks he neu ali y o scien i ic heo ies. Pu nam (2002) poin s ou ha among
non-epis emic and non-cogni i e concep s, especially among e hical ones, he e a e
alues ha can be conside ed bo h no ma i e and desc ip i e depending on he pe -
spec i e adop ed. These a e cases like ‘c uel’, ‘di y’ o he like. He calls hem ‘ hick’
concep s and uses hem p ecisely o p o ide coun e examples o he ac / alue dis-
inc ion o , in ou con ex , non-epis emic alue/epis emic alue.6I he use o e ms
such as hese in e wining ac s and alues is una oidable in scien i ic p ac ice and
easoning, any p oposal o scien i ic hypo heses and esul s can ha dly be alue- ee.
This seems o unde mine any hesis ad oca ing he impa iali y o science and he
neu ali y o i s alues.7
[4]Some philosophe s o science a gue ha non-epis emic alues ha e an in-
dispensable legi ima e ole o play in assessing he isks in ol ed in he accep ance
o scien i ic hypo heses and decisions ( he a gumen om induc i e isk) (Ellio &
Richa ds 2017; Ellio & S eel 2017, Pa III; B own 2013; Douglas 2000, 2009b,
p.58; Rolin 2015, p.161; S eel 2010; Wilhol 2009; see Rudne 1953, p.2). One o i s
p emises is ha he accep ance o ejec ion o hypo heses ( he usual ask o scien-
is s) in ol es unce ain y. In doing so, a scien is has o decide whe he he a ailable
e idence is s ong enough o wa an accep ance. This decision depends on he isks
in ol ed. I scien is s accep a alse hypo hesis, he e may be a cos associa ed wi h
his ype o e o . I hey ejec a ue hypo hesis, he cos will be o a di e en kind.
The key p emise o his c i ique is ha he assessmen o he cos s in ol ed in bo h
ypes o e o s is a ma e o mo al alue judgemen (Rudne 1953, p.3).
As can be seen, i is assumed ha de ending he ole o non-epis emic and non-
cogni i e alues in a emp ing scien i ic hypo heses o heo ies does no necessa ily
ule ou he possibili y o c ea ing egula o y amewo ks in which bene icial and
ha m ul consequences o such alues could be dis inguished (C espo 2019; Diek-
PRINCIPIA 28(4): 517–540 (2024)
522 Juan Bau is a Bengoe xea
mann & Pe e son 2013). This makes i possible o p opose al e na i es, some imes
qui e complex, o he ideal acco ding o which science is ee o alues. Those would be
al e na i es ha should ede ine and dis inc ly loca e se e al epis emic no ions such
as eliabili y, objec i i y, and a ionali y (Douglas 2009b, Chap e 6; Koskinen 2021;
Ca w igh 2022).8These epis emic alues guide he ask pe o med oge he wi h
cogni i e alues, including he explana o y powe (Ca w igh e al. 2022; P i cha d
2021, p.5523; In emann 2015; see Psillos 2015; Koskinen & Rolin 2022, p.193). The
combined o m o all he p oduc s de i ed om aking hese alues in o conside a-
ion would be ha o a ne wo k. The loca ion o alues, in u n, would no be gi en
by he classical dis inc ion be ween epis emic and cogni i e e sus social and e hical
alues, bu a he by he ole played by each ype o each oken: di ec o indi ec
(Douglas 2009b, p.95–7). Bu we poin ou he ele ance o he ‘ oken’ p ecisely be-
cause, con a y o wha Douglas claims, we conside ha he ole o he explana o y
powe o scien i ic heo ies (unde s ood as a cogni i e alue) canno be labelled as
‘indi ec ’ ou cou . Bo h mechanis ic explana ions and (ma hema ic-like) s uc u al
explana ions in expe imen al linguis ics will se e as an illus a i e case o ou c i-
ique o Douglas.
In o de o loca e and seek o unde s and his c i ique, we mus see ha he
accep abili y o scien i ic ac i i y would he e o e be based on i s eliabili y, ela ed o
he use o e idence, oge he wi h he assessmen o he lack o bias, he signi icance
o scien i ic heo ies and all hei p oduc s, hei comple eness, hei consis ency and
o he cogni i e alues. The cons an conside a ion o hese epis emic and cogni i e
alues in scien i ic ac i i y is no exemp om he pa icipa ion o social and e hical
alues as well. The ne wo k in which hey play a pa is complex and angled, and i is
he e o e use ul o un a el, a leas pa ially, hose phases o he scien i ic en e p ise
whe e he use o alues is mo e o less di ec . The complex na u e o scien i ic ac i i y
should be emphasised i we we e o unde s and i s p oduc s in any o he phases in
which i akes place. Unde s anding his kind o complexi y is linked o he eliabili y
we expec om science om he e y momen we begin o de elop i (Ca w igh e
al. 2022; B igand 2015; Po e 2006; Lacey 1999, p.38–9).
3. Values in o scien i ic esea ch undamen al phases
Values a e inhe en in scien i ic p ac ice aimed a p o iding eliable knowledge
(McIn y e 2019, p.230; Douglas 2009b, p.113–4; Po e 2006, p.78).9How o a ic-
ula e such eliabili y olled in alues, including social and e hical ones, is a p essing
goal o he philosophy o science (see Ca w igh e al. 2022, Pa II, 4). I alues,
he e o e, canno be e used om scien i ic ac i i y, hen eliabili y o , as i may be,
app oxima ion o u h, would ha e o be ob ained h ough di e en s a egies in
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 523
scien i ic esea ch.10 One consis s o ecognizing he plu ali y o alues a he di e -
en s ages o esea ch, om he e y beginning, when esea che s selec hei goals
and wo king me hods— heo y-guessing, making conjec u es o hypo heses, and so
on (Vilhalemm 2016; Poppe 1994, p.95))—, o he inal implemen a ion s age. In
all hese s ages, he weigh o alues can be es ima ed in some way, and hei linkage
o p ope ly epis emic s a egies based on e idence and da a can be calib a ed.
I is in e es ing o c i ically examine he oles ha alues can play in scien i ic
ac i i ies (Lacey 2011; Lacey & Ma iconda 2014).11 This examina ion s a s om he
unde s anding ha scien i ic ac i i y can be s uc u ed in phases, speci ically in ou
(’Ph’ onwa ds) (see Douglas 2009b, p.88–9): [Ph-1]Adop ion and de elopmen o a
esea ch s a egy; [Ph-2]ga he ing da a; [Ph-3]cogni i e and/o epis emic e alua-
ion o he heo ies, models o hypo heses p oposed o sol e some p oblem iden i ied
in he ini ial adop ion; and [Ph-4]implemen a ion o he ob ained scien i ic knowl-
edge.
[Ph-1]Adop ion and de elopmen o a esea ch s a egy: The i s decision con-
ce ns he esea ch s a egy; he esea ch opic mus be decided. One usual s a ing
poin o a esea ch is o p opose some conjec u e, hypo hesis, p oblem o heo y
(Ca w igh 2022, p.19–20). Tha poin se es o selec and shape he phenomena
ha scien is s in end o in es iga e. In o de o do his, esea che s i emediably ake
non-cogni i e alues in o accoun . Adop ing a esea ch s a egy is simila o p opose
a complex amewo k o scien i ic inqui y made up o a ious elemen s, among which
he iden i ica ion o an ini ial p oblem and o a wo king guide hypo hesis helping o
sol e he o me a e wo ema kable ones (Lacey & Ma iconda 2014, p.647). This
ini ial guide o scien i ic inqui y is c ucial in o de o es ablish he ele an ypes o
empi ical da a, he app op ia e desc ip i e ca ego ies o making obse a ional e-
po s, as well as he heo ies linked o da a. The link be ween da a and heo ies, i
mus be emphasized, equi es he adop ion o a s a egy om he beginning.
[Ph-2]Ga he ing da a: Once he objec o esea ch has been de e mined, scien is s
en e he jus i ica ion phase, in which da a and e idence a e collec ed and accumu-
la ed,12 a emp s a e made o jus i y ce ain hypo heses, o explain ac s in he ield
s udied o , in some cases, o p edic hose ac s. The ways o collec ing e idence can
be e y a ied i we compa e he na u al sciences wi h he social sciences. In he la -
e , i is e y common o appeal o in e iews, epo s, ield wo k and, in gene al, o
ools ha a e o eign o mos o he na u al sciences (Douglas 2009b, p.88).
[Ph-3]Epis emic e alua ion o heo ies (models, hypo heses): Once he necessa y
e o s ha e been made o con i m and e i y (o disp o e o alsi y) ce ain hypo he-
ses, he s onges o he bes suppo ed hypo hesis is chosen. To con i m a hypo hesis
is o ind e idence o easons o i and no agains i , so ha he hypo hesis likeli-
hood inc eases. Howe e , a his s age, con a y o wha Douglas (2009b) ad oca es,
cogni i e alues can ac as c i e ia o epis emic e alua ion and be independen om
PRINCIPIA 28(4): 517–540 (2024)
524 Juan Bau is a Bengoe xea
non-cogni i e alues (McIn y e 2019, Landes 2020). As heo ies o hypo heses end
o be accep ed o e u ed mo e clea ly acco ding o some c i e ion o eliabili y, a
heo y will be accep ed o a ield o phenomena i i exhibi s i s own cogni i e al-
ues (scope, accu acy, explana o y powe , p edic i e powe , e c.) no shown by any
al e na i e heo y—o shown by he la e in a lesse deg ee. This is always made
in he ligh o empi ical da a and he ele an and su icien e idence collec ed om
obse a ion, expe imen a ion o e en simula ions, and always on he basis o he
used heo y and o he phenomena o be s udied. Non-cogni i e alues and ce ain
non-empi ical (me aphysical, specula i e) con ic ions play no ole in he ‘Ph-3’ o
he heo y assessmen choice. This kind o alues ha dly, o no a all, de e mines he
in es iga ion in Ph-3 (see Be z 2013, p.212).
[Ph-4]The implemen a ion o knowledge: To he ex en ha a hypo hesis o heo y
is consolida ed in a scien i ic discipline, and depending on esea che s’ in e es s and
alues, he conclusions and he p ac ical, applicable and e en egula o y op ions a e
analyzed and selec ed o decision-making in science policy (see Bengoe xea & Tod
2021, p.57–62; Bengoe xea 2024, p.113). This is also called ‘ he phase o implemen-
a ion’ o science (see González 2013). I may adop he o m o a ansi ion om he
na u al sciences’ con ex o he con ex o social sciences ( om biochemis y o eg-
ula ion, ypically). By applying scien i ic knowledge, some ideals a e p ojec ed and
esea che s se e ce ain in e es s ha e lec speci ic non-cogni i e alues. Science
is applied because scien is s bo h expec some bene i s s imula ed by hei ini ial e-
sea ch in e es s and assume o o e come some po en ial nega i e consequences o
scien i ic applica ions (Lacey & Ma iconda 2014).
I is no a necessa y condi ion ha hese ou phases de elop in a successi e
o de ( o example, Ph-2 and Ph-3 could in luence Ph-1, o he p ocesses in Ph-4
could (allegedly) a ec p e ious phases). Since he p ocesses o ob aining scien i ic
knowledge and i s applica ion a e subjec o a ying deg ees o unce ain y, science is
assumed o be allible ( his is also a condi ion o doing science oday). I means ha
scien is s some imes ha e o make decisions (o hypo heses) unde isky condi ions
(Ellio & Richa ds 2017, p.2). In doing so, i is mos usual o hink abou he (also
social) in e es s o he cogni i e en e p ise, as well as he possible e o s o such de-
cisions. Thinking abou consequences equi es aking in o accoun social, poli ical,
e hical and economic alues, all o which can in luence he p ocess o e i ica ion
and jus i ica ion o a hypo hesis by he scien i ic communi y. The e o e, i seems o
us ha he Ph-1/Ph-4 scheme is a good ool ha could make easie he ep esen a-
ion o he cogni i e p ocesses es ablished and de eloped in science. The case s udy o
s uc u al explana ion (sec ion 4) is p oposed o suppo his hypo hesis. In he case
o expe imen al linguis ics, s uc u al explana o y echniques in ol e in oducing a
cogni i e alue (explana o y powe ) in Ph-2 and Ph-3 as a undamen al elemen . This
does no p eclude he need o eso o non-cogni i e alues in some cases. In his
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 525
sense, as we app oach he social o human sciences, whe e human beha io is he
objec o esea ch, i is mo e di icul o a oid he in luence o non-cogni i e alues.
Human language is a e y complex phenomenon wi h many edges. The e o e, he
a icula ion and e alua ion o e idence and hypo heses u ns ou o be e y complex
and allows o di e en in e p e a ions ha may be mo i a ed by mul iple alues o
in e es s, a leas in p inciple, al hough he main goal is hei eliabili y as an epis-
emic alue ha condi ions scien i ic ac i i y. Bu explana o y powe —concei ed as
a cogni i e alue playing a di ec ole in Ph-2 and Ph-3—is loca ed in a complex ne -
wo k along wi h o he alues—cogni i e, e hical, and social—and can help de elop
a mo e eliable and mo e in eg al science (Douglas 2009b, p.95), also in he case o
expe imen al linguis ics.
4. Complex ne wo ks, s uc u al explana ions and he s udy
o some ea u es o language
Hempel’s nomological-deduc i e model o explana ion ceased o be he quin essen-
ial app oach in disciplines such as biology and neu osciences s udying li ing and
cogni i e sys ems (Ki che 1989; Douglas 2010; Hempel 1965; see Diez; Khali a;
Leu idan 2013; Douglas 2009a, p.449; Salmon 1989, p.94–101; F iedman 1974).
Ins ead o seeking uni e sally applicable laws o p edic i e models, esea che s in o
hese ields ha e sough o unde s and he beha io o li ing o ganisms by analyz-
ing hem in pa s (s uc u es). This explana ion s a egy was based on he s udy o
mechanisms (Machame ; Da den; C a e 2000; Mo eno & Suá ez 2020, p.146).
Howe e , mechanis ic explana ion is a ype o explana ion ha canno always
accoun o ce ain aspec s o he biology o o ganisms (de elopmen , ep oduc ion,
e c.) (Alle a; Díez; Fede ico 2017) no deal wi h o he p ope ies o complex sys ems
ha a e be e explained by ools om complexi y sciences (Huneman 2018, B ig-
and ; G een; O’Malley 2017). Some hing can be said o be complex i i is made up o
in e connec ed pa s and i he unde s anding o i s beha io equi es comp ehend-
ing no only he beha io o i s pa s, bu also he way in which hey mu ually ac o
gi e ise o an o e all beha io . Global desc ip ion equi es desc ibing he pa s, and
hese in u n a e desc ibed in e ms o he whole (Ba -Yam 1997, p.1).
The mos impo an cha ac e is ics o be aken in o accoun o complex sys ems
can be summa ized as ollows (Ba -Yam 1997, p.5): (1) he elemen s o en i ies o a
sys em (and hei quan i y), (2) he in e ac ions among hem (and hei s eng h),
(3) hei o ma ion and unc ion, (4) hei di e si y, (5) hei en i onmen , and (6)
hei ac i i ies. Bu wha is complexi y? To de ail a desc ip ion o a complex sys em, i
is use ul o unde s and how he concep o complexi y is ela ed o ha o eme gence
(Ba -Yam 1997, p.6).
PRINCIPIA 28(4): 517–540 (2024)
532 Juan Bau is a Bengoe xea
di ec ole in accep ing o ejec ing a hypo hesis (Douglas 2009b, p.110). She claims
ha nei he cogni i e no social alues a e a good eason by hemsel es o accep
o no accep some heo y. We pa ially disag ee wi h his s a emen . The explana-
o y powe o a linguis ic heo y is one elemen among o he s (always combined
wi h o he elemen s (see Ca w igh e al. 2022)) ha some imes plays only an indi-
ec ole (guiding and suppo ing decision-making by esea che s o egula o y sci-
en is s),18 bu a o he imes can ac di ec ly because scien i ic explana o y powe ,
unlike o he cogni i e alues, is based on he de elopmen o explana ions hem-
sel es, which is di ec ly linked o wo king e idence and da a. This no mally occu s
in phases Ph-2 and Ph-3, al hough i is ela ed o he o he phases. Wha we do no
demand he e is ha he explana ion has o be based on mechanisms, in gene al, no
on causal mechanisms, in pa icula . A s uc u al explana ion, based on ma hema i-
cal opology, can p o ide a ep esen a ion o an explanandum ha e lec s a ne wo k
and ha can be added o help sol e se e al p oblems.
5. Conclusion
The openness o philosophy o o he ields o in e es akes he o m o a ocus on
issues o bo h social and scien i ic in e es . To his end, i is essen ial o wo k hand
in hand wi h scien is s hemsel es (Douglas 2010, p.317). Some philosophe s ha e
claimed ha philosophy o science should egula ly ge engaged in o scien i ic p ac-
ices in o de o unde s and hei unde lying philosophical issues (see Ki che 2023,
p.16). I is abou a kind o social ele ance ha we migh desc ibe as ’being philo-
sophically a en i e’ o hose aspec s o science ha a e pa icula ly impo an o
socie y.
Gi en he deside a um ha philosophy should ob ain g ea e ele ance among
scien is s and he gene al public (Ki che 2023, p.16), we ha e conside ed one nexus
connec ing jus he h ee domains o alues. The alues—whe he epis emic, cogni-
i e, social, o e hical—we e ini ially concei ed (acco ding o he ideal o a alue- ee
science) as an undesi able in luence on he scien i ic en e p ise and on i s in eg i y,
objec i i y, and eliabili y. Howe e , hese concep s hemsel es al eady ook he o m
o alues, a leas epis emic and cogni i e ones. A space was made o hem, bu no
o he supposedly ha m ul ‘social and e hical alues’. This image was soon supe -
seded in he philosophy o science by a se ies o well-es ablished c i ical a gumen s,
o igina ed by Rudne in 1953, so ha mo e de ailed and complex app oaches, such
as ha o Hea he Douglas o Hugh Lacey, among o he s, es ablished phases o sci-
ence in which he in eg a ion o alues g adually a ied. Thanks o his new sha ing
o he complex p ocedu es o making heo ies and p ac ices in science, i has been
possible o unde s and ha he ole o alues is loca ed in some combined p ocesses
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 533
simila o wha Neu a h mean wi h his ’Ballung concep s’. In hese ci cums ances,
Douglas u he inco po a ed he in e es ing idea ha alues, o wha e e kind, may
in luence a ious phases o science, bu will ne e play a di ec ole (Douglas 2010,
p.328; 2009b, p.96, 112).
Well, his is p ecisely he hesis ha we c i icize he e. The cogni i e alue ‘ex-
plana o y powe ’ is essen ial o suppo he epis emic alue ‘ eliabili y’. And scien i ic
explana ion is no limi ed o playing an indi ec guiding ole in decision making, bu ,
in he sense o he Ballung men ioned abo e, i is de eloped oge he wi h he han-
dling o da a, e idence and o he p ope ly scien i ic p oduc s and means. We ha e
ied o show his in he case o expe imen al linguis ics. We ha e p oposed he no-
ion o s uc u al explana ion, based on ma hema ics, which se es as a complemen
o many esul s o mechanis ic explana ions, since s uc u al explana ions a e no
concei ed as me e ‘aids’ o decision making in he cen al phases o science (Ph-2
and Ph-3, o example).
La e we ha e mo ed owa ds hose h ee deside a a h ough he applica ion o
a s uc u al explana o y app oach. The ask o elabo a ing a no ion o scien i ic ex-
plana ion is p ope ly philosophical and, in o de o ca y i ou , we ha e also aken
in o accoun an app oach based on a di ec ole o cogni i e alues. We ha e p e-
sen ed a case om expe imen al linguis ics wi h which we wo k in close p oximi y
o alues and e idences linked o language acquisi ion. We hink ha his is a su i-
cien demons a ion, a he momen , o posi i ely alue he scien i ic (linguis ic) and
social ad ances o an expe imen al science in con inuous p og ess and helped by a
philosophy o science, epis emologically oo ed, ha a emp s o adequa ely concep-
ualize i .
Re e ences
Abbuhl, R.; Gass, S.; Mackey, A. 2013. Expe imen al esea ch design. In R. Podes a; D.
Sha ma (ed.), Resea ch Me hods in Linguis ics, p.116–134. Camb idge: Camb idge Uni-
e si y P ess.
Alle a, K.; Díez, J.; Fede ico, L. 2017. Models, heo y s uc u e and mechanisms in biochem-
is y: The case o allos e ism. S udies in His o y and Philosophy o Biology and Biomedical
Sciences 63: 1–14.
An ony, L. 1993. Quine as eminis : The adical impo o na u alized epis emology. In L.M.
An ony; C.E. Wi (ed.), A Mind o One’s Own: Feminis Essays on Reason and Objec i i y,
p.110–153. Boulde : Wes iew.
Ba -Yam, Y. 1997. Dynamics o Complex Sys ems. Reading, MA.: Addison-Wesley. Ba abási,
A.-L. 2002. Linked: The New Science o Ne wo ks. Camb idge, MA: Pe seus.
Ba celó-Coblijn, L.; Co ominas-Mu a, B.; Gomila, A. 2012. Syn ac ic ees and small-wo ld
ne wo ks: syn ac ic de elopmen as a dynamical p ocess. Adap i e Beha io 20(6): 427–
442.
PRINCIPIA 28(4): 517–540 (2024)
534 Juan Bau is a Bengoe xea
Ba celó-Coblijn, L.; Duguine, M.; I u zun, A. 2019. The Eme gence o Hubs in Complex
Syn ac ic Ne wo ks and he DP Hypo hesis: The Rele ance o a Linguis ic Analysis. In À.
Massip; G. Bel-Enguix; A. Bas a das-Boada (ed.), Complexi y Applica ions in Language and
Communica ion Sciences, p.273–288. London: Sp inge .
Ba celó-Coblijn, L.; Se na, D.; Isaza, G.; Cas illo, L.F.; Bedia, M. 2017. Ne lang: A so wa e
o he linguis ic analysis o co po a by means o complex ne wo ks. PlosOne 12(8): 1–15.
Ba ke , G., Ki che , P. 2014. Philosophy o Science: A New In oduc ion. Ox o d: Ox o d Uni-
e si y P ess.
Bech el, W.; Richa dson, R. C. 1993. Disco e ing complexi y: Decomposi ion and localiza ion
as s a egies in scien i ic esea ch. P ince on: P ince on Uni e si y P ess.
Bengoe xea, J. B. 2024. Impa cialidad y dema cación de alo es en la ac i idad cien í ica.
Re is a Ibe oame icana de Ciencia y Tecnología 19(55): 107–125.
Bengoe xea, J. B.; Tod , O. 2021. Decision-Making in he Nu i ion Sciences: A C i ical Anal-
ysis o Scien i ic E idence o Assessing Heal h Claims. Manusc i o 44(3): 42–69.
Be z, G. 2013. In de ence o he alue ee ideal. Eu opean Jou nal o Philosophy o Science
3: 207–220.
B igand , I. 2015. Social alues in luence he adequacy condi ions o scien i ic heo ies: be-
yond induc i e isk. Canadian Jou nal o Philosophy 45: 326–356.
h ps://doi.o g/10.1080/00455091.2015.1079004.
B igand , J.; G een, S.; O’Malley, M. A. 2017. Sys ems biology and mechanis ic explana ion.
In S. Glennan; P. Illa i (ed.), The Rou ledge handbook o mechanisms and mechanical
philosophy, p.362–374. London: Rou ledge.
B own, M. J. 2013. Values in Science beyond Unde de e mina ion and Induc i e Risk. Phi-
losophy o Science 80(5): 829–839.
Bueno, O. 2011. Pa ial T u h and Visual E idence. P incipia 15(2): 249–270.
Ca w igh , N. 2022. A Philosophe Looks a Science. Camb idge: Camb idge Uni e si y P ess.
Ca w igh , N.; Ha die, J. 2012. E idence-based policy: a p ac ical guide o doing i be e . New
Yo k: Ox o d Uni e si y P ess.
Ca w igh , N.; Ha die, J.; Mon uschi, E.; Soleiman, M.; Th eshe , A.C. (ed.). 2022. The Tangle
o Science: Reliabili y Beyond Me hod, Rigou , and Objec i i y. Ox o d: Ox o d Uni e si y
P ess.
Chomsky, N.; Mille , G.A. 1963. In oduc ion o he o mal analysis o na u al languages. In L.
Duncan; R.R. Bush; E. Galan e (ed.), Handbook o ma hema ical psychology, p.269–321.
New Yo k: John Wiley.
Co ominas-Mu a, B. 2007. Ne wo k s a is ics on ea ly English Syn ax: S uc u al c i e ia.
a Xi e-p in .h ps://a xi .o g/abs/0704.3708. Access 12/12/2022.
Co ominas-Mu a, B.; Val e de, S.; Solé, R.V. 2009. The on ogeny o scale- ee syn ax ne -
wo ks: Phase ansi ions in ea ly language acquisi ion. Ad ances in Complex Sys ems (ACS)
12: 371–392.
Coy e, K. Z.; Schlu e , J.; Fos e , K. R. 2015. The ecology o he mic obiome: Ne wo ks, com-
pe i ion, and s abili y. Science 350(6261): 663–666.
C espo, R.F. 2019. Libe al Na u alism and Non-epis emic Values. Founda ions o Science 24:
247–273.
Deulo eu, R.; Suá ez, J.; Pé ez-Ce e a, A. 2021. Explaining he beha iou o andom ecolog-
ical ne wo ks: The s abili y o he mic obiome as a case o in eg a i e plu alism. Syn hese
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 535
198(3): 2003–2025.
Diekmann, S.; Pe e son, M. 2013. The Role o Non-Epis emic Values in Enginee ing Models.
Sci. Eng. E hics 19: 207–218.
Díez, J. A.; Khali a, K.; Leu idan, B. 2013. Gene al heo ies o explana ion: Buye s bewa e.
Syn hese 190: 379–396.
Douglas, H. E. 2000. Induc i e Risk and Values in Science. Philosophy o Science 67(4): 559–
579.
Douglas, H. E. 2009a. Rein oducing P edic ion in Explana ion. Philosophy o Science 79:
444–463.
Douglas, H. E. 2009b. Science, Policy, and he Value-F ee Ideal. Pi sbu gh: Uni e si y o Pi s-
bu gh P ess.
Douglas, H. E. 2010. Engagemen o P og ess: Applied Philosophy o Science in Con ex .
Syn hese 177(3): 317–335.
Douglas, H. E. 2014. Values in Social Science. In N. Ca w igh ; E. Mon uschi (ed.), Philosophy
o Social Science: A New In oduc ion, p.162–182. Ox o d: Ox o d Uni e si y P ess.
Dup é, J. 2007. Fac and alue. In H. Kincaid; J. Dup é; A. Wylie (ed.), Value-F ee Science:
Ideals o Illusions, p.27–41. Ox o d: Ox o d Uni e si y P ess.
Ellio , K. C.; Richa ds, T. (ed.). 2017. Explo ing Induc i e Risk: Case S udies o Values in Sci-
ence. Ox o d: Ox o d Uni e si y P ess.
Ellio , K. C.; S eel, D. (ed.). 2017. Cu en Con o e sies in Values and Science. New Yo k:
Rou ledge.
F iedman, M. 1974. Explana ion and scien i ic unde s anding. Jou nal o Philosophy 71: 5–19.
Glennan, S. 1996. Mechanisms and he na u e o causa ion. E kenn nis 44(1): 49–71.
González, W.J. 2013. Value Ladenness and he Value-F ee Ideal in Scien i ic Resea ch. In C.
Lue ge (ed.), Handbook o he Philosophical Founda ions o Business E hics, p.1503–1521.
Sp inge : Do d ech .
G een, S.; Jones, N. 2016. Cons ain -based easoning o sea ch and explana ion: S a egies
o unde s anding a ia ion and pa e ns in biology. Dialec ica 70(3): 343–374.
G ies, S. T.; Newman, J. 2013. C ea ing and using co po a. In R.J. Podes a; D. Sha ma (ed.),
Resea ch Me hods in Linguis ics, p.257–287. Camb idge: Camb idge Uni e si y P ess.
Haack, S. 1998. Science as Social? –Yes and No. In Mani es o o a Passiona e Mode a e, p.104–
122. Chicago: The Uni e si y o Chicago P ess.
Haack, S. 2016. Se ious Philosophy. Spazio Filoso ico 18: 395–407.
Hacking, I. 1983. Rep esen ing and In e ening: In oduc o y Topics in he Philosophy o Na u al
Science. Camb idge: Camb idge Uni e si y P ess.
Ha ding, S. 1986. The Science Ques ion in Feminism. New Yo k: Co nell Uni e si y P ess.
Hempel, C. 1965. Aspec s o scien i ic explana ion and o he essays in he philosophy o science.
New Yo k: F ee P ess.
Hooke , C. 2011. In oduc ion o Philosophy o Complex Sys ems. In C. Hooke (ed.), Philos-
ophy o Complex Sys ems, p.841–909. Ams e dam: Else ie .
Huneman, P. 2010. Topological explana ions and obus ness in biological sciences. Syn hese
177: 213–245.
Huneman, P. 2018. Ou lines o a heo y o s uc u al explana ion. Philosophical S udies 175(3):
665–702.
PRINCIPIA 28(4): 517–540 (2024)
536 Juan Bau is a Bengoe xea
In emann, K. 2015. Dis inguishing be ween legi ima e and illegi ima e alues in clima e mod-
eling. Eu opean Jou nal o Philosophy o Science 5: 217–232.
Kelle , E. F. 2005a. The cen u y beyond he gene. Jou nal o Biosciences 30: 3–10.
Kelle , E. F. 2005b. Re isi ing ‘scale- ee’ ne wo ks. BioEssays 27: 1060–1068.
Kelle , S. H.; Longino, H. E.; Wa e s, C.K. 2006. Scien i ic Plu alism. Minneapolis: Uni e si y
o Minneso a P ess.
Kincaid, H.; Dup é, J.; Wylie, A. (ed.). 2007. Value-F ee Science? Ideals and Illusions. Ox o d:
Ox o d Uni e si y P ess.
Ki che , P. 1989. Explana o y uni ica ion and he causal s uc u e o he wo ld. In P. Ki che ;
W. Salmon (ed.), Scien i ic explana ion, p.410–505. Minneapolis: Uni e si y o Minneso a
P ess.
Ki che , P. 2023. Wha ’s he Use o Philosophy?. Ox o d: Ox o d Uni e si y P ess.
Koskinen, I. 2021. Objec i i y in con ex s: wi hholding epis emic judgemen as a s a egy o
mi iga ing collec i e bias. Syn hese 199: 211–225.
Koskinen, I.; Rolin, K. 2022. Dis inguishing be ween legi ima e and illegi ima e oles o al-
ues in ansdisciplina y esea ch. S udies in His o y and Philosophy o Science 91: 191–198.
Lacey, H. 1997. The Cons i u i e Value o Science. P incipia 1(1): 3–40.
Lacey, H. 1999. Is Science Value-F ee? Values and Scien i ic Unde s anding. London: Rou ledge.
Lacey, H. 2004. Is The e a Signi ican Dis inc ion be ween Cogni i e and Social Values? In P.
Machame ; G. Wol e s (ed.), Science, Values, and Objec i i y, p.24–51. Pi sbu gh: Uni e -
si y o Pi sbu gh P ess.
Lacey, H. 2011. A impa cialidade da ciência e as esponsabilidades dos cien is as. Scien iæS-
udia 9(3): 487–500.
Lacey, H.; Ma iconda P.R. 2014. O modelo das in e ações en e as a i idades cien í icas e os
alo es. Scien iæS udia 12(4): 643–668.
Landes, J. 2020. The a ie y o e idence hesis and i s independence o deg ees o Indepen-
dence. Syn hese 198: 10611–10641.
Laudan, L. 1984. Science and Values: The Aims o Science and Thei Role in Scien i ic Deba e.
Be keley: Uni e si y o Cali o nia P ess.
Longino, H. 1990. Science as Social Knowledge. Camb idge, MA: P ince on Uni e si y P ess.
Longino, H. 1996. Cogni i e and Non-Cogni i e Values in Science: Re hinking he Dicho omy.
In L. H. Nelson; J. Nelson (ed.), Feminism, Science, and The Philosophy o Science, p.39–58.
Do d ech : Kluwe .
Machame , P.; Da den, L.; C a e , C. 2000. Thinking abou mechanisms. Philosophy o Science
67: 1–25.
MacWhinney, B.; Snow, C. 1990. The Child Language Da a Exchange Sys em: an upda e.
Jou nal o Child Language 17(2): 457–472.
Mahne , M. 2022. Na u alismo. T ans. F ancisco José Mo a Po eda [Na u alismus]. Pamplona:
Lae oli.
Ma ins Dos Reis, C.R., De Pa a Pilla , V. 2018. Valo es, es a égias de pesquisa e aplicação
do conhecimen o: os Campos Sulinos en ques ão. P incipia 22(3): 461–483.
McIn y e, L. 2019. The Scien i ic A i ude: De ending Science om Denial, F aud, and Pseudo-
science. Camb idge, MA: The MIT P ess.
McMullin, E. 1982. Values in Science. In P.D. Asqui h; T. Nickles (ed.), PSA: P oceedings o
he 1982 Biennial Mee ing o he Philosophy o Science Associa ion (Volume Two, Symposia
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 537
and In i ed Pape s), p.3–28. Eas Lansing, MI: Philosophy o Science Associa ion.
McNama a, T. 2006. Validi y and alues: In e ences and gene alizabili y in language es -
ing. In M. Chalhoub-De ille; C.A. Chapelle; P. Du (ed.), In e ence and Gene alizabili y in
Applied Linguis ics, p.27–45. Ams e dam: John Benjamins.
Mo eno, Á.; Suá ez, J. 2020. Plu ali y o explana o y s a egies in biology: mechanisms and
ne wo ks. In W.J. González (ed.), Me hodological P ospec s o Scien i ic Resea ch: F om
P agma ism o Plu alism, p.141–165. Cham: Sp inge .
Ninio, A. 2006. Language and he leaning cu e: A new heo y o syn ac ic de elopmen . Ox o d:
Ox o d Uni e si y P ess.
Poppe , K. R. 1994. The My h o he F amewo k. London: Rou ledge.
Po e , E. 2006. Feminism and Philosophy o Science: An In oduc ion. London: Rou ledge.
Pou na i, M. 2008. The Dis inc ion Be ween Epis emic and Non-Epis emic Values in he Na -
u al Sciences. Science & Educa ion 17: 669–676.
P i cha d, D. 2021. In ellec ual i ues and he epis emic alue o u h. Syn hese 198: 5515–
5528.
Psillos, S. 2015. E idence: wan ed, ali e o dead. Canadian Jou nal o Philosophy 45: 357–
381.
Pu nam, H. 2002. The Collapse o he Fac /Value Dicho omy and O he Essays. Camb idge, MA:
Ha a d Uni e si y P ess.
Radde , H. 2003. Technology and Theo y in Expe imen al Science. In The Philosophy o Sci-
en i ic Expe imen a ion, p.152–173. Pi sbu gh: Uni e si y o Pi sbu gh P ess.
Resnik, D. B.; Ellio , K.C. 2023. Science, Values, and he New Dema ca ion P oblem. Jou -
nal o Gene al Philosophy o Science 54: 259–286. h ps://doi.o g/10.1007/s10838-022-
09633-2
Rezaee, H. S.; Bika aan-Behesh , H. 2023. Value-F ee Ideal is an Epis emic Ideal: An Objec ion
o he A gumen om Induc i e Risk. P incipia 21(1): 137–163.
Risjo d, M. 2023. Philosophy o Social Science: A Con empo a y In oduc ion. 2nd Edi ion. New
Yo k: Rou ledge.
Rolin, K.H. 2015. Values in Science: The Case o Scien i ic Collabo a ion. Philosophy o Science
82(2): 157–177.
Rudne , R. 1953. The Scien is Qua Scien is Makes Value Judgmen s. Philosophy o Science
20(1): 1–6.
Salmon, W. 1989. Fou Decades o Scien i ic Explana ion. Minneapolis: Uni e si y o Minneso a
P ess.
Seoane, L.; Solé, R. 2018. The mo phospace o language ne wo ks. Scien i ic Repo s 8: 10465.
Sh ade -F eche e, K. 1993. Bu ying Unce ain y: Risk and he Case Agains Geological Disposal
o Nuclea Was e. Be keley: Uni e si y o Cali o nia P ess.
S eel, D. 2010. Epis emic Values and he A gumen om Induc i e Risk. Philosophy o Science
77: 11–34.
Thu ne , S.; Hanel, R.; Klimek, P. 2018. In oduc ion o he Theo y o Complex Sys ems. Ox o d:
Ox o d Uni e si y P ess.
Tomasello, M. 2003. Cons uc ing a language: A usage-based heo y o language acquisi ion.
Camb idge, MA: Ha a d Uni e si y P ess.
Van F aassen, B.C. 1980. The Scien i ic Image. Ox o d: Ox o d Uni e si y P ess.
PRINCIPIA 28(4): 517–540 (2024)
538 Juan Bau is a Bengoe xea
Vilhalemm, R. 2016. Chemis y and he p oblem o plu alism in science: an analysis conce n-
ing philosophical and scien i ic disag eemen s. Founda ions o Chemis y 18: 91–102.
Wilhol , T. 2009. Bias and alues in scien i ic esea ch. S udies in His o y and Philosophy o
Science 40(1): 92–101.
Zuidema, W.; de Boe . B. 2013. Modelling in he language sciences. In R.J. Podes a; S. De-
yani (ed.), Resea ch Me hods in Linguis ics, p.428–445. Camb idge: Camb idge Uni e -
si y P ess.
No es
1On he mul iple lines o esea ch adop able in medicine, as a ield o high scien i ic and
social in e es , see o example Ba ke and Ki che 2014, p.153–4.
2Ad oca es o he ideal o a alue- ee science migh objec ha he in luence o alues, i
any, is always p io o any esea ch. This is he seminal idea o Reichenbach (1938) when he
p oposed he dis inc ion be ween he con ex o disco e y and ha o jus i ica ion, acco ding
o which, he mission o he scien is would be o seek and ind his jus i ica ion, o es ablish
a hypo hesis well, o jus i y i , co obo a e i and e i y i in such a way ha non-epis emic
alues do no in e ene in hese p ocedu es. This is p ecisely wha is no accep ed ou cou
oday.
3Al hough we ollow he ou -phase pa e n, close o Douglas and Lacey, o he au ho s,
as is he case o Ma ins Dos Reis & De Pa a Pilla (2018, p.464–8), analyze he oles ha
alues exe in i e phases o scien i ic ac i i y: adop ion o he esea ch s a egy, de elopmen
o ha s a egy, cogni i e e alua ion o hypo heses and heo ies, sp ead o scien i ic esul s,
and applica ion o scien i ic knowledge.
4Douglas (2009b, p.93) men ions as cogni i e alues simplici y, scope, consis ency, p edic-
i e accu acy, ecundi y and— he one ha in e es s us mos he e—explana o y powe . Lacey
(1999, p.58–60), on he o he hand, speaks o empi ical adequacy, explana o y and uni y-
ing powe , encapsula ion o possibili ies, in e nal consis ency, consonance, suppo be ween
heo ies, and he capaci y o sol e puzzles.
5Fo example, he ac ha some indi iduals die due o he side e ec s o a d ug w ongly
conside ed sa e, o ha o he indi iduals die om a disease because hey did no ha e access
o a ea men w ongly conside ed unsa e.
6Pu nam’s hick concep s a e qui e simila o Neu a h-Ca w igh ’s Ballung concep s, al-
hough he o me a e speci ically placed in he e hical deba e.
7John Dup é (2007) openly claims ha hick e hical e ms canno be elimina ed om
science, a leas om ce ain pa s o i . Scien i ic hypo heses, heo ies and esul s conce n us
jus because hey a ec ou human in e es s, which makes i necessa y o exp ess hem wi h a
ocabula y lecked wi h hick e hical concep s. And while i is some imes possible o ansla e
‘ hickness’ in o ‘impa iali y’ and ‘neu ali y’, he consequence o his always esul s in some
loss o ce ain human in e es s, since he e a e always, mo e o less su ep i iously, in e es s
o his kind a s ake. “I mos o all o physics is alue ee, i is no because physics is science
bu because mos o physics simply doesn’ ma e o us”, Dup é (2007, p.31) p o oca i ely
says.
PRINCIPIA 28(4): 517–540 (2024)
Ne wo ks, Explana ions and Values in Linguis ics 539
8Anyway, his pe spec i e is no unanimous in he philosophy o science. Be z (2013),
o example, demands mo e p ecision in he ea men o he ole o (non-epis emic) alues
in he con ex o a c i icism o a science wi hou alues. Acco ding o him, his c i icism is
an amoun o jus logging a dead ho se (see P i cha d 2021, Pou na i 2008).
9As McIn y e (2019, p.230) w i es, “while i is ue ha we should no le ou hopes,
wishes, belie s, and ‘ alues’ colou ou inqui y in o he ‘ ac s’ abou human beha iou , his
does no mean ha alues a e unimpo an . Indeed, as i u ns ou , ou commi men o he
scien i ic a i ude is an essen ial alue in conduc ing scien i ic inqui y”.
10Such a s a egy would be he explana o y one, o example, aimed a ob aining i s own
esul s hanks o “a dense in e wo en ne o scien i ic cons uc ions ha cons ain and suppo
each o he (...) ideas, concep s, heo ies, expe imen s, measu es, middle-le el p inciples,
models, me hods o in e ence, esea ch adi ions, da a, na a i es” (Ca w igh 2022, p.12,
20; see also Ca w igh e al. 2022), and so on. Among hem, we conclude ha also alues
(cogni i e, epis emic and non-epis emic) mus be in oduced.
11We ollow he s anda d o ma o he ou phases and lea e ou one phase ha Lacey and
Ma iconda (2014, p.645) do ake in o accoun : sp eading scien i ic esul s. We hink ha his
and ‘ou ’ Ph-4 can be uni ied. In any case, his does no a ec he goal o ou a icle.
12Mahne (2022, p.68) quali ies he sligh di e ence be ween he concep o da a and ha
o e idence. Acco ding o him, da a a e he esul (exp essed linguis ically) o an empi ical
ope a ion (obse a ion, measu emen , expe imen ), bu hey a e no e idences. As such, da a
do no say any hing by hemsel es. A da um dbecomes e idence only in ela ion o a hypo h-
esis ho heo y . This is impo an in many epis emological con ex s, bu he e we will ea
bo h concep s equi alen ly because he sligh di e ence does no in e e e wi h ou wo king
hesis.
13Fo cases o weake ela ionships such as isomo phism o homomo phism, see Bueno
(2011, p.253–4).
14Deulo eu, Suá ez and Pé ez-Ce e a (2021, p.2010–1) p esen a s uc u al, no mecha-
nis ic, explana ion o he s abili y beha io o he mic obiome. They poin ou ha , al hough
Coy e¸ Schlu e and Fos e (2015) appea o p o ide a model o he mechanism de ined in
e ms o en i ies, ac i i ies, and hei o ganiza ion, he de ails o he causal pa hway ha
cause he s able beha io a e ac ually dilu ed in he complexi y o a ma hema ical analysis.
15In ma hema ics and physics, a small-wo ld ne wo k is a kind o ne wo k o which mos
o he nodes a e no neighbo s o each o he , al hough, ne e heless, almos all o hem
can be eached om any sou ce node h ough a ela i ely sho numbe o s eps be ween
hem (Ba abási 2002, p.53). Tha is, he cons uc ion o a small-wo ld ne wo k depends
on (1) he exis ence o a small-wo ld phenomenon as such (any wo nodes in he ne wo k
communica e o e a ela i ely small in e media e node pa h (small numbe o nodes), since
he maximum dis ance be ween wo nodes g ows loga i hmically along wi h he o al numbe
o nodes in he ne wo k), and (2) ha i possesses high clus e ing coe icien s o , in o he
wo ds, he indica ion ha , i wo nodes a e no di ec ly connec ed o each o he , he e is
a high p obabili y ha hey connec h ough he in e en ion o o he nodes. The e o e, i
is a p ope y o a small wo ld ha i p esen s ne wo ks in which, despi e he exis ence o a
la ge numbe o nodes, i is possible o ind sho communica ion pa hs be ween hem. On he
o he hand, a scale- ee ne wo k is a speci ic kind o complex ne wo k. In a scale- ee ne wo k,
some nodes a e highly connec ed, i.e., hey ha e a la ge numbe o links o o he nodes,
PRINCIPIA 28(4): 517–540 (2024)
540 Juan Bau is a Bengoe xea
al hough he deg ee o connec ion o almos all o he nodes is qui e low. Ba abási (2002,
p.58) poin s ou ha he Web, o example, does no ha e a dis ibu ion o he usual deg ee
o connec i i y. Ins ead, a ew nodes, which a e called hubs, a e much mo e connec ed han
he es . An example is he case o ne wo ks o neu ons in o ganisms wi h ne ous sys ems.
In hese, i is qui e no mal o use e y equen ly a ac ion o he neu ons while lea ing mos
o he es o he neu ons ba ely u ilized.
16CHILDES (Child Language Da a Exchange Sys em) is a co pus (as o 2015 i had con en s–
ansc ip s, audio, and ideo—in 26 languages om 230 di e en co po a) o se e as a
cen al eposi o y o da a o i s language acquisi ion. See MacWhinney and Snow (1990).
17A SAN so wa e is a high-speed local ne wo k ha con ains mul iple se s o block s o age
de ices. Among some o i s ea u es, SAN can consolida e s o age de ices (eases manage-
men ), e icien ly pools s o age de ices, imp o es pe o mance by educing s o age la ency,
imp o es applica ion a ailabili y h ough edundancy, simpli ies da a backup (no se e in-
e ac ion), and suppo s mul iple kinds o s o age de ices.
18Douglas (2009b, p.108–12) analyzes he indi ec ole o alues (cogni i e and social) in
he case o die hyls ilbes ol (DES) (a compound ha ope a ed as an es ogen). Acco ding
o he , he explana o y alue in his case only plays an o ien ing ole in decision making.
Howe e , as Coy e, Schlu e and Fos e (2015), Deulo eu, Suá ez and Pé ez-Ce e a (2021),
and Mo eno and Suá ez (2022) poin ou , he e a e o he cases in which he explana ion
adop s mo e complex o ms, as in he case o he mic odiome. We hink ha hese mo e
complex explana ions, which a e some imes o a s uc u al-ma hema ical na u e o , also,
in eg a ed wi h a mechanis ic explana ion o ma (Deulo eu; Suá ez; Pé ez-Ce e a 2021,
p.2022), do play a di ec ole, as we ha e ied o show in he case o expe imen al linguis ics.
Acknowledgmen s
G an PID2020-113449GB-I00, unded by MICIU/AEI/10.13039/501100011033/, and g an
PID2023-147251NB-I00 unded by MICIU/AEI/10.13039/501100011033 and by ERDF/EU.
I would also like o hank he wo anonymous e e ees o hei con ibu ion o he imp o e-
men o his pape .
PRINCIPIA 28(4): 517–540 (2024)
