T acing back he his o y o I alian
A ibu i e-Apposi i e Noun+Noun compounds:
Fi s ou comes
Jan Radimský (Uni e si y o Sou h Bohemia, České Budějo ice)
ABSTRACT
The pape p esen s he ini ial indings o a esea ch p ojec ha in es iga es he o igins o I alian
ATAP NNs, such as pa olaN chia eN (keywo d) o luogoN simboloN (place.symbol— symbolic place).
This s udy is based on ame iculously chosen sample o 1,924 NN ypes ha include 47 di e en mod-
i ie s as hei igh mos cons i uen , sou ced om Google n-g ams equency lis s (2020), which
ep esen he mos ex ensi e diach onic linguis ic da a cu en ly a ailable. While ecen li e a u e
sugges s ha I alian ATAP NNs eme ged in he la e hal o he 20 h cen u y and gained signi ican
p oduc i i y only a e 2000, ou da a indica e ha his p ocess occu ed app oxima ely 50 yea s
ea lie , pa icula ly o ano able subse o N2-based semi-schema ic cons uc ions. Wi h espec
o he heo e ical amewo ks o Cons uc ion Mo phology, Rela ional Mo phology, and Diach onic
Cons uc ion G amma , he analysis o he da a p o ides an insigh in o he in e ac ion be ween
he ully schema ic ATAP NN cons uc ion and a ious semi-schema ic cons uc ions (N2-based
amilies). Mo eo e , he s udy a emp s o iden i y key diach onic s ages in he de elopmen o he
pa e n.
KEYWORDS
ATAP, a ibu i e-apposi i e compounds, compounding, diach ony, I alian
DOI
h ps://doi.o g/10.14712/18059635.2023.2.3
1 INTRODUCTION
I alian Noun+Noun compounds (NN compounds, hence o h) ha e been ex en-
si ely in es iga ed om asynch onic poin o iew (see, o an o e iew, Radim-
ský 2015) due o hei aising p oduc i i y and he wide a ie y o pa e ns a es ed
in con empo a y I alian. In pa icula , A ibu i e-Apposi i e NNs (ATAP, hence-
o h), such as pa olaN chia e.N (keywo d) o luogoN simboloN (place.symbol— sym-
bolic place), which ha e been eme ging a a as pace especially in he pas wo de-
cades, aised ali ely deba e conce ning hei heo e ical s a us and classi ica ion
(see e e ences in Sec ion 2.1.), as well as di e en mo phosyn ac ic p ope ies,
including, namely, in lec ion o lexical inse ion (Radimský 2018, G andi, Nissim
and Tambu ini 2011). Howe e , much less a en ion has been paid o he diach ony
o NN compounds, which seem o ep esen a ela i ely ecen inno a ion in Ro-
mance. In o de o ill his gap, his a icle p esen s i s ou comes o a esea ch ha
aims o ace back he o igin and de elopmen o I alian ATAP NNs based on Google
n-g ams equency lis s (2020), which a e he mos ex ensi e diach onic linguis-
ic da a cu en ly a ailable. The obse a ions p esen ed below d aw on aca e ully
OPEN
ACCESS
JAN RADIMSKý 137
selec ed sample o 1,924 ATAP NN compounds ha con ain 47 di e en modi ie s
as hei igh mos cons i uen , wi h equency da a co e ing he pe iod om 1800
o 2019.
The pape is s uc u ed as ollows: Sec ion 2 ou lines basic p ope ies o I al-
ian ATAP NNs and summa ises exis ing knowledge abou hei diach onic de el-
opmen . Sec ion 3 b ie ly in oduces he unde lying heo e ical backg ound ha
s ems om Cons uc ion Mo phology (Booij 2010, 2016), Rela ional Mo phology
(Jackendo and Aud ing 2020), and Diach onic Cons uc ion G amma (Hilpe
2015, T augo and T ousdale 2013, Goldbe g 2019, Hilpe 2021, among o he s).
Sec ion 4 desc ibes he da a-ga he ing p ocess, while Sec ion 5 p esen s he dia-
ch onic p o ile o he comple e ATAP sample and he diach onic p o ile o selec ed
N2-based amilies o “semi-schema ic cons uc ions”. Finally, Sec ion 6 sum-
ma ises he esul s.
2 ATTRIBUTIVE-APPOSITIVE NN COMPOUNDS
2.1 DELIMITATION OF ATAP COMPOUNDS
The g oup o ATAP compounds, concei ed in he in luen ial classi ica ion o com-
pounds by Bise o and Scalise (2005; 2009) as one o he h ee majo compounding
pa e ns, along wi h coo dina e and subo dina e compounds, ha e gi en ise o s im-
ula ing discussions (c . Ba oni, Gue a a and Pi elli 2009; G andi 2009; G andi, Nis-
sim and Tambu ini 2011; A codia, G andi and Mon e mini 2009) ha poin o he
ac ha his class o compounds is pa icula ly ha d o de ine on o mal g ounds. In
sho , Scalise and Bise o (2009) a gue ha ATAP compounds ea u e an a ibu i e
ela ionship be ween he head and i s modi ie , he la e exp essing a“p ope y”
o “quali y” o he head, bu hey do no p o ide clea c i e ia on how o es i in he
case o NN s uc u es, besides he ac ha he modi ie “is o en o be in e p e ed
me apho ically” (Scalise and Bise o 2009: 51–52). Wi hou en e ing in o de ails o
his deba e (see Radimský 2015: 92–112 o ade ailed o e iew), we will delimi he e
he g oup o I alian ATAP NNs on o mal g ounds using he ans o ma ional es in-
oduced in Radimský (2015: 92–102, see also A naud 2003: 12–13 o F adin 2009 o
F ench), as ou lined in he ollowing pa ag aphs.
The es is based on awidely sha ed assump ion (c . Sal i and Vanelli 2011) ha
he a ibu i e ela ionship ypically exis s be ween aquali ying adjec i e and he
co esponding head noun, and ha his ela ionship can be consis en ly pa aph ased
using ap edica i e exp ession wi h acopula e b. Since nouns may also ea u e as
subjec complemen s1 o copula e bs, he same es can be used o de ec an a ibu-
i e ela ionship in NN s uc u es. The e o e, ATAP NNs wi h an abs ac modi ie
may be ans o med in ap edica i e exp ession di ec ly (1), while ATAP NNs wi h
aconc e e noun used as ame apho ic modi ie may be ans o med a leas indi ec ly,
using an ope a o such as I alian come (“like a”) (2).
1 In I alian: complemen o p edica i o del sogge o.
OPEN
ACCESS
138 LINGUISTICA PRAGENSIA 2/2023
(1) luogo simbolo → ques o luogo è un simbolo
[place]N [symbol]N
symbolic place → his place is asymbol
(2) pesce palla → ques o pesce è come una palla
[ ish]N [globe]N
globe ish → his ish is like aglobe
No ice ha he es yields ung amma ical o inapp op ia e ou pu s o subo dina e
NN compounds (3) as well as o NA ph ases wi h a ela ional adjec i e (4), which is
p ecisely wha makes i sui able o se apa ATAP NN s uc u es (1, 2) om subo -
dina e NNs (3).2
(3) p o ezione pe sone → *ques a p o ezione è pe sone
[p o ec ion]N [pe sons]N
p o ec ion o pe sons → * his p o ec ion is pe sons
(4) sis ema ne oso → *ques o sis ema è ne oso
[sys em]N [ne ous]A
ne ous sys em → * his sys em is ne ous
Amo e ine-g ained classi ica ion o I alian ATAP NNs, in oduced in Radimský
(2015: 158–159), u he se s apa a ibu i e NNs wi h an abs ac modi ie (1) om
apposi i e NNs wi h aconc e e modi ie (2), since only in he la e g oup modi ie s
ha e ame apho ical in e p e a ion. Howe e , as clea as his di e ence may seem a
i s sigh , i s empi ical applica ion comes up agains he ac ha — especially in he
case o nominal modi ie s wi h highe ype equencies in ATAP compounds— he
me apho ical in e p e a ion o he noun is o en well “lexicalised” e en wi hou be-
ing limi ed o he NN pa e n. Fo ins ance, he monolingual Zinga elli (2011) dic ion-
a y lis s among he basic meanings o he wo ds chia e (key) and lampo (ligh ning,
lash) hei me apho ical in e p e a ions “impo an ” (5) and “sho / quick” (6), e-
spec i ely, desc ibing hese in e p e a ions as common e en ou side o he NN pa -
e n, as he examples wi hin (5–6) show.
(5) chia e 2— Elemen o, da o ope sona di impo anza i ale pe comp ende e, in e p-
e a e, isol e e qlco. […] Quel pe sonaggio è la chia e di u a la icenda.
chia e 2— Elemen , ac o pe son o i al impo ance o unde s anding, in-
e p e ing, esol ing s h. […] [Example:] Tha cha ac e is he key o he whole
a ai .
2 Fo he sake o comple eness, i should be added ha his ans o ma ional es also yields
posi i e esul s o endocen ic coo dina e NN compounds, such as la o a o e-s uden e
(s uden wo ke ), in which bo h componen s deno e simila concep s si ua ed on he same
hie a chy le el (see Radimský 2015: 102–112 o ade ailed discussion). These NNs a e no
included wi hin he ATAP class o NNs he e.
OPEN
ACCESS
JAN RADIMSKý 139
(6) lampo 3— Ciò che ha b e issima du a a: la gio inezza è un lampo.
lampo 3 — Wha is e y sho -li ed: [Example:] you h is a lash.
lampo 4— Pe sona, animale o eicolo mol o eloce: quel cane è un lampo.
lampo 4— Ve y as pe son, animal o ehicle: [Example:] ha dog is a lash.
In hese cases, i would be apu ely academic ques ion o ask whe he NNs such as pa-
ola chia e (keywo d) o eno lampo ( as ain, li . “ lash ain”) a e a he apposi i e
(since he in e p e a ion o he modi ie is me apho ical wi h espec o he p ima y
conc e e meaning o ha noun) o a ibu i e (since he in e p e a ion o he modi-
ie is based on one o i s meanings which happens o be me apho ical). To pu i di -
e en ly, he me apho ical in e p e a ion o modi ie s, such as (5–6), is no bound o
he NN pa e n. Tha is why no u he dis inc ion will be made be ween a ibu i e
and apposi i e NN compounds in his a icle.
Finally, awo d is in o de abou he ac ha some au ho s sugges ha selec ed
modi ie s in ATAP compounds (i.e., such as chia e o lampo in he examples men-
ioned abo e) should a he be analysed as “noun-clad” in a iable adjec i es (G andi,
Nissim and Tambu ini 2011, G andi 2009, bu see Tho n on 2004: 528–530 o Radim-
ský 2015: 22–29). Since hese elemen s a e p ima ily nouns, wi h selec ed syn ac ic
(bu no mo phological) ea u es ypical o quali ica i e adjec i es, he deba e seems
inconclusi e on o mal g ounds. We hus conside ha he choice o classi ying hese
modi ie s combined wi h head nouns as “ATAP NN compounds” is one o he wo pos-
sible me hodological op ions, whose pa icula ad an age is ha i makes i possible
o desc ibe ATAP NNs as an in eg al pa o I alian NNs ha ha e en e ed aphase o
dynamic e olu ion in he pas cen u y.3
2.2 HISTORY OF ATAP COMPOUNDS
As Raine (2021) poin s ou , he cu en knowledge abou he e olu ion and mode n
p oli e a ion o Romance NN compounds is s ill ex emely limi ed. The pa e n does
no display any con inui y om La in compounding, bu a he s ems om a a ie y
o he e ogeneous syn ac ic cons uc ions whose numbe seems e y low in I alian,
a leas un il he end o he 19 h cen u y.
In an ex ensi e diach onic s udy o I alian compounds based on he Codi co -
pus o li e a y I alian, Micheli (2020a: 91–93) ound only 3 ATAP NNs in Old I alian
(7a–c) and 15 ATAP NNs in Middle I alian (Micheli 2020a: 145, 152–155); he la e
e e o animals— especially ishes— (8a), plan s, ege ables, o o he conc e e
subs ances (8b).
3 On he o he hand, le us add ha some speci ic egula NN s uc u es we e delibe a e-
ly kep apa om NN compounds. These in ol e especially cases whe e he modi ie is
ap ope noun, i.e. a e e en ial exp ession, such as so ella Ma ia (sis e Ma y), ha we
conside as ins ances o egula syn ax. O he s in ol e sel - e e en ial modi ie s, such
as l’agge i o “buono” ( he adjec i e “buono”) o il nume o qua o ( he numbe ou ). Fo
amo e de ailed discussion see Radimský (2015:40–44).
OPEN
ACCESS
140 LINGUISTICA PRAGENSIA 2/2023
(7a) pescespada— swo d ish ( ish+swo d)
(7b) pescepo co— g ey igge ish ( ish+pig)
(7c) a camensa— la ge cupboa d (a k+ able)
(8a) pescecane (dog ish, sha k, ish+dog), pesceca allo (ho se ace loach, ish+ho se),
pescesega (saw ish, ish+saw), pesceluna (ocean sun ish, ish+moon), o mica-
leone (an -lion, an +lion), g illo alpa (mole c icke , c icke +mole)
(8b) mela ancia (swee o ange, apple+o ange), ca ol io e (cauli lowe , cabbage+
lowe ), ca ol apa (kohl abi, cabbage+ u nip), cascio io e ( ype o cheese
simila o peco ino, cheese+ lowe ), ca as accio (looking like was epape ,
pape +sh edded), e baspada (aga e, g ass+swo d), ce alacca (sealing wax,
wax+ ype o esin), gommalacca (shellac, gum+ ype o esin)
Un o una ely, he da a ex ac ion in he s udy by Micheli (2020a) conside s only
compounds w i en in he “ igh ” o “hyphena ed” o m. I compounds in he “loose”
o m (i.e., w i en as wo sepa a e wo ds) we e also ex ac ed, a ew o he examples
migh be ound in he Codi co pus o comple e he sample ( o ins ance, edizione p in-
cipe— edi io p inceps). Ne e heless, his s ill indica es an ex emely limi ed p es-
ence o ATAP NNs in he li e a y language be o e he mid-20 h cen u y.
Though he da a om he Codi co pus (Micheli 2022) do no go u he han he
yea 1947, based on seconda y sou ces— o , a he , he lack o eliable in o ma ion
in seconda y sou ces— Micheli (2020b: 120) assumes ha he ATAP pa e n has
eached eal p oduc i i y and dissemina ion only since he 21s cen u y.
The da a analysed in his a icle aim o shed mo e ligh on he e olu ion o he
I alian ATAP NNs in he pas wo cen u ies.
3 THEORETICAL FRAMEWORK
As Cons uc ion Mo phology (Booij 2010; 2016) and Rela ional Mo phology (Jack-
endo and Aud ing 2020) a e usage-based models, schemas a ailable in he Con-
s uc icon cap u e gene aliza ions o e ac i ical mass o al eady a es ed wo ds.
“Cons uc ionaliza ion” mus he e o e be based on p e ious indi idual “Inno-
a ion” (T augo and T ousdale 2013). The p ocess o cons uc ionaliza ion may
p oduce acomplex hie a chical ne wo k o schemas ha cap u e di e en le els
o gene aliza ion in he Cons uc icon. In he case o ATAP NN compounds, such
ane wo k may be schema ised in (9a-d), showing in u n indi idual ins ances o
ATAP NNs (9d), semi-schema ic cons uc ions based on a epea ed componen (9c),
he ully schema ic ATAP NN schema (9b) and he mos gene al le -headed NN pa -
e n (9a).
(9a) Le -headed NN cons uc ion
[NiNj]Nk ↔ [Ni-head Nj-non-head]k
OPEN
ACCESS

JAN RADIMSKý 141
(9b) ATAP NN cons uc ion
[NiNj]Nk ↔ [Ni-head È (COME) UN(A) Nj-non-head]k
[NiNj]Nk ↔ [Ni-head IS (LIKE) ANj-non-head]k
(9c) ATAP NN semi-schema ic cons uc ions
c.1 SUB NN semi-schema ic cons uc ions based on he same modi ie (N2)
[Ni chia e]Nk ↔ [Ni-head È chia e/impo an e]k
[Ni chia e]Nk ↔ [Ni-head IS key/impo an ]k
c.2 SUB NN semi-schema ic cons uc ions based on he same head (N1)
[pesce Nj]Nk ↔ [pesce che È COME UN(A) Nj-non-head]k
[pesce Nj]Nk ↔ [ ish ha IS LIKE ANj-non-head]k
(9d) Indi idual ins ances o NNs
d.1 pa ola chia e (keywo d)
d.2 pesce spada (swo d ish)
d.3 uolo chia e (key ole)
d.4 pesce palla (globe ish)
As udy ha aims a desc ibing and accoun ing o he eme gence o acompounding
wo d- o ma ion pa e n wi hin his heo e ical pe spec i e has o ask he ollowing
ques ions, namely:
Fi s , i is necessa y o iden i y in he diach onic da a p ima y lexical inno a ions
(“leading wo ds” o “a ac o s”, see e.g., Ha mann 2019: 316) ha may be subjec o
subsequen e ymological analyses. Such leading wo ds appea in he ea ly s ages o
he de elopmen o pa e ns, whe e hey a e expec ed o ha e highe oken equen-
cies, hough la e , hei oken equency may dec ease signi ican ly. E en hough
o he in e es ing “a ac o s” wi h high ype equencies may also eme ge in subse-
quen s ages, hei o igin is po en ially di e en : ea ly a ac o s canno esul om
he applica ion o p oduc i e schemas, so ha hei e ymology mus be s udied and
accoun ed o indi idually.
Second, p og essi ely eme ging highe -o de cons uc ions ha e o be de ec ed
and desc ibed. Recen s udies on compounding emphasise he p ominen ole o
lowe -o de schemas ha equi e an in-dep h in es iga ion (see Hilpe 2015;
Ha mann 2019). Indeed, he same gene al p ope ies o highe -o de cons uc-
ions may co espond o e y di e en con igu a ions o lowe -le el cons uc-
ions: ahigh ype equency o agene al pa e n, such as (9a) o (9b), may be due
ei he o an inc easing numbe o lowe -le el cons uc ions (9c) o a he o ahigh
ype equency o jus a ew lowe -le el cons uc ions (9c). In he la e case, he
highe -o de cons uc ions do no ha e su icien co e age (Goldbe g 2019: 51–73),
which en ails ha he inc easing ype equency o lowe -o de cons uc ions
does no s eng hen he men al ep esen a ion o highe -o de cons uc ions (Hil-
pe 2015; 2019).
This a icle will ocus on he la e ques ion om aquan i a i e poin o iew.
On asample o 1,924 ATAP NNs wi h 47 di e en N2s (modi ie s), we will analyse
OPEN
ACCESS
142 LINGUISTICA PRAGENSIA 2/2023
p og essi ely eme ging highe -o de cons uc ions (9b) and (9c), as well as hei in-
e ac ion in ime.
In o de o iden i y diach onic ends and d aw eg ession lines, Theil-Sen es i-
ma o will be used and supplemen ed wi h he Mann-Kendall es o signi icance
es ing (Py hon implemen a ion by Hussain and Mahmud 2019). These ank-based
non-pa ame ic me hods a e sui able o es any o m o dependence (no only lin-
ea ), hey do no assume ano mal dis ibu ion o e o s, and hey a e no sensible
o ou lie s, which makes hem pa icula ly sui able o end iden i ica ion o wo d
usage in diach onic co po a (Ko ář and He man 2013).
4 DATA
The esea ch is based on ex ensi e diach onic da a d awn om he Google books
co pus in he o m o aw equency lis s ha ha e been published as he 3 d e -
sion o I alian Google n-g ams.4 The size o he unde lying Google books co pus is
120,410,089,963 okens. Da a o he ex ac ion o NN compounds come om p e-
ea ed big ams and ig ams o cap u e compounds wi h space-sepa a ed and hy-
phen-sepa a ed componen s, espec i ely.5 The n-g ams we e il e ed and me ged
oge he in o asingle da ase labeled “i 2020_bi”6 using he p ocedu e desc ibed
in de ail by Radimský (2022); he whole i 2020_bi da ase ( om which unc ion
wo ds ha e been il e ed ou ) comp ises 19,319,372 non-lemma ised ypes. F om
his lis , only o ms ha po en ially ma ch he o ms o I alian nouns acco ding
o he Mo phI dic iona y (Zanche a and Ba oni, 2005) we e il e ed ou , which
yielded alis o 2,723,566 NN candida es. As e en his educed lis s ill con ains an
ex emely high a e o alse posi i es, u he il e ing o ATAP NNs had o be done
manually.
ATAP compounds a e known o clus e a ound modi ie -based amilies: some au-
ho s e en assume ha nominal modi ie s in ATAP compounds cons i u e a ela i ely
limi ed se o nouns (Ba oni, Gue a a, and Pi elli 2009). Tha is why hei manual
ex ac ion consis ed in iden i ying po en ial modi ie s on he igh mos posi ion o
NNs ollowed by sys ema ic manual checking o all co esponding heads on he le -
mos posi ion.
Po en ial nominal modi ie s (hence o h N2s, by i ue o hei igh mos posi-
ion in he N1N2 s uc u e unde in es iga ion) we e iden i ied in wo ways. Fi s ,
N2s ha appea ed in he sample o 1,800 p esen -day I alian ATAP NNs om Rad-
imský (2015) we e pu on he lis . Second, he lis was comple ed by u he N2s
ex ac ed manually om he Zinga elli (2011) dic iona y ha ma ks such nouns
ei he by aspeci ic seconda y pa -o -speech ca ego y labeled “in a iable adjec i e
4 h ps://s o age.googleapis.com/books/ng ams/books/da ase s 3.h ml
5 Compounds w i en as one g aphical wo d ha e no been conside ed as his o hog aphy
is ypical only o neoclassical igh -headed compounds in mode n I alian (see Radimský
2015: 49–50).
6 The i 2020-bi da ase is a ailable o download a : h ps://os .io/46qcd/
OPEN
ACCESS
JAN RADIMSKý 143
unc ion” (in unzione di agg. in .), o by ano e explaining ha hey may be used as
non-head nouns in NN s uc u es (pospos o aun sos . / pospos o al sos .). The dic-
iona y da a was added o ensu e ha some p ominen olde modi ie s ha a e no
longe in use we e no missing om he sample. The cu en lis o N2 candida es
con ains 319 non-lemma ised wo d o ms ( hough u he il e ing may show ha
many o hem a e alse posi i es).
In he subsequen manual iden i ica ion o head nouns (N1s), we ocused p omi-
nen ly ( hough no exclusi ely) on N2s wi h highe ype equencies. Up o his poin ,
i was possible o check and il e whole amilies o head nouns (N1s) ha co espond
o 47 di e en N2s, hei comple e lis including espec i e ype equencies is gi en
in Appendix. In o de o achie e he highes accu acy du ing da a ex ac ion, he ol-
lowing measu es we e aken.
a) The majo i y o compounds we e manually e i ied h ough Google Books. Ap e-
limina y e iew o asample consis ing o 10–20 con ex ualised ins ances ( okens)
o en allows o dis inguishing be ween au hen ic ATAP NN compounds and alse
posi i es. Fac o s con ibu ing o he p esence o alse posi i es on he candida e
lis include e o s a ising om op ical cha ac e ecogni ion (OCR) and complex-
i ies in mo phological o syn ac ic ambigui y. This e i ica ion p ocess success-
ully helped elimina e nume ous alse posi i es. Howe e , he equency coun s
o he iden i ied ue posi i e ypes may s ill con ain some noise, which is un-
a oidable due o he es ic ed access o o iginal Google Books da a and he shee
olume o da a in ol ed.
b) Wo d o ms we e used as basic uni s ins ead o lemmas, as i has been obse ed
ha due o mo phological ambigui y, many in lec ed o ms o en con ibu e o
alse posi i es in ac ual ex s. As mos modi ie s in I alian ATAP NNs a e in a i-
able (Radimský 2015: 49–50; 2018), he wo d/lemma dis inc ion has a ela i ely
limi ed ele ance in he case o N2s.
The cu en sample comp ises 1,924 ATAP NNs, unde s ood as non-lemma ised ypes,
ea u ing 47 di e en N2s (modi ie s) and 1,148 di e en N1s (head nouns). Fo each
compound ( ype), da ed numbe s o occu ences in Google books a e a ailable om
1800 o he p esen wi h ayea -by-yea p ecision.
5 RESULTS
5.1 DIACHRONIC PROFILE OF THE SCHEMATIC ATAP PATTERN
The diach onic e olu ion o ATAP compounds can be analyzed on a ious le els o ab-
s ac ion, namely, subs an ial cons uc ions (indi idual ypes), semi-schema ic con-
s uc ions (N1- o N2-based amilies), and he ully schema ic ATAP cons uc ion. As
ou lined in Sec ion 3, we will ocus on he la e wo cases. Le us i s adop he bi d’s
eye iew by making aglobal o e iew o he whole sample o 1,924 ATAP NNs ha
co esponds o he schema ic cons uc ion (10).
OPEN
ACCESS
144 LINGUISTICA PRAGENSIA 2/2023
(10) ATAP NN cons uc ion
[NiNj]Nk ↔ [Ni-head È (COME) UN(A) Nj-non-head]k
[NiNj]Nk ↔ [Ni-head IS (LIKE) ANj-non-head]k
Figu e 1 gi es adiach onic o e iew o he numbe o ypes ( ype equency) o he
ATAP pa e n om he 1850s o 2019. Since he size o he unde lying co pus a ies
in diach ony, ype equency is exp essed in ela i e e ms as he numbe o ypes
di ided by he co pus size in he espec i e yea , which co esponds o he classical
p oduc i i y measu e e e ed o as Realised P oduc i i y by Baayen (2009). The e-
sul was mul iplied by he cons an 108 so ha i in ui i ely app oaches he o de o
magni ude o he o iginal ype equency da a.7 The ac ha Realised P oduc i i y is
es ic ed o “pas achie emen ”, usually quo ed as ad awback o his measu e when
applied o synch onic co po a, does no ep esen an issue when diach onic da a a e
a ailable.
The cu e in Figu e 1 shows a ela i ely s eady inc ease in he ype equency om
he beginning o he 20 h cen u y on, wi h wo pe iods o s eepe slope in 1950–1980
and 2000–2019, espec i ely.
Ne e heless, such an o e all pic u e does no show how he lowe -le el amilies
con ibu e o he inc ease, i.e., whe he he gene al inc ease in ype equency is due
o an inc easing numbe o many di e en lowe -le el cons uc ions o a he o
ahigh ype equency o jus a ew lowe -le el cons uc ions. Indeed, he N2-based
amily wi h N2 chia e (key), which is he one ha has by a he highes ype e-
quency in he sample, also has an inc easing ealised p oduc i i y since he 1930s, as
7 Fo ins ance, in he yea 2019, he absolu e ype equency 1,603 co esponds o he ela-
i e ype equency 1,331.16 displayed in Figu e 1.
Figu e 1:Realised p oduc i i y o ATAP NNs in diach ony
(V/N×108, whe e Vis he numbe o ypes, and N is he co pus size in he espec i e yea )
OPEN
ACCESS
JAN RADIMSKý 151
o alse posi i es which is impossible o elimina e— o men ion a leas he mos
impo an ones. Ne e heless, Google n-g am da a s ill ep esen he iches and
mos accu a e sou ce o he diach onic analysis o w i en language o he pas
wo cen u ies cu en ly a ailable.
REFERENCES
A codia, G.F., N.G andi and F.Mon e mini
(2009) Hie a chical NN compounds in
ac oss-linguis ic pe spec i e. I alian Jou nal o
Linguis ics 21/1, 11–33.
A naud, Pie e J.L. (2003) Les composés imb e-
pos e (T a aux du C.R.T.T.). Lyon: P esses Uni .
de Lyon.
Baayen, R.H. (2009) Co pus linguis ics in
mo phology: Mo phological p oduc i i y. In:
Lüdeling, A. and M.Ky ö (eds) Handbooks o
Linguis ics and Communica ion Science, 899–919.
Be lin / New Yo k: Mou on de G uy e . doi:10.
1515/9783110213881.2.899.
Ba oni, M., E.Gue a a and V.Pi elli (2009)
Sulla ipologia dei compos i N+N in
i aliano : p incipi ca ego iali ed e idenza
dis ibuzionale acon on o. In: Fe a i, G.,
R.Bena i and M.Mosca (eds), Linguis ica
emodelli ecnologici di ice ca : a i del XL
Cong esso in e nazionale di s udi della Socie à
di linguis ica i aliana (SLI) : Ve celli, 21–23
se emb e 2006, 73–96. Roma: Bulzoni Edi o e.
Bise o, A. and S.Scalise (2005) The
Classi ica ion o Compounds. Lingue
elinguaggio 2, 319. doi:10.1418/20728.
Booij, G. (2010) Cons uc ion mo phology (Ox o d
linguis ics). Ox o d / New Yo k: Ox o d
Uni e si y P ess.
Booij, G. (2016) Cons uc ion Mo phology.
In: Hippisley, A. and G.S ump (eds) The
Camb idge Handbook o Mo phology, 424–448.
1.edn. Camb idge: Camb idge Uni e si y
P ess. doi:10.1017/9781139814720.016.
F adin, B. (2009) IE, Romance: F ench. In:
Liebe , R. and P.Š ekaue (eds) The Ox o d
Handbook o Compounding, 417–435. Ox o d:
Ox o d Uni e si y P ess. doi:10.1093/
ox o dhb/9780199695720.013.0022.
Goldbe g, A.E. (2019) Explain me his: c ea i i y,
compe i ion, and he pa ial p oduc i i y o
cons uc ions. P ince on, New Je sey: P ince on
Uni e si y P ess.
G andi, N. (2009) When Mo phology ’Feeds’
Syn ax: Rema ks on Noun > Adjec i e
Con e sion in I alian Apposi i e Compounds.
In: Mon e mini, F., G.Boyé and J.Tseng (eds)
Selec ed P oceedings o he 6 h Décemb e es,
112–124. Some ille (MA): Cascadilla
P oceedings P ojec .
G andi, N., M.Nissim and F.Tambu ini (2011)
Noun-clad adjec i es. On he adjec i al s a us
o non-head cons i uen s o I alian a ibu i e
compounds. Lingue eLinguaggio 10/1, 161–176.
doi:10.1418/34543.
Ha mann, S. (2019) Compound wo lds
and me apho landscapes: A ixoids,
allos uc ions, and highe -o de
gene aliza ions. Wo d S uc u e 12/3, 297–333.
doi:10.3366/wo d.2019.0151.
Hilpe , M. (2015) F om hand-ca ed
o compu e -based: Noun-pa iciple
compounding and he upwa d s eng hening
hypo hesis. Cogni i e Linguis ics 26/1, 113–147.
doi:10.1515/cog-2014-0001.
Hilpe , M. (2019) Highe -o de schemas in
mo phology: Wha hey a e, how hey wo k,
and whe e o ind hem. Wo d S uc u e 12/3,
261–273. doi:10.3366/wo d.2019.0149.
Hilpe , M. (2021) Ten lec u es on diach onic
cons uc ion g amma (Dis inguished lec u es
in cogni i e linguis ics ol.26). Leiden /
Bos on: B ill.
Hussain, Md. and I.Mahmud (2019)
pyMannKendall: apy hon package o non
pa ame ic Mann Kendall amily o end
es s. Jou nal o Open Sou ce So wa e 4/39.
1556. doi:10.21105/joss.01556.
Jackendo , R. and J.Aud ing (2020) The
ex u e o he lexicon: ela ional mo phology
and he pa allel a chi ec u e. Fi s edi ion.
OPEN
ACCESS

152 LINGUISTICA PRAGENSIA 2/2023
Ox o d / New Yo k: Ox o d Uni e si y
P ess.
He man, O. and V.Ko ář (2013) Me hods o
De ec ion o Wo d Usage o e Time. In:
Ho ák, A. and P.Rychlý (eds), P oceedings o
Recen Ad ances in Sla onic Na u al Language
P ocessing, RASLAN 2013, 79–85. B no: T ibun
EU.
Micheli, M.S. (2020a) La o mazione delle
pa ole: i aliano eal e lingue (S udi supe io i;
Linguis ica 1237). 1a edizione. Roma: Ca occi
edi o e.
Micheli, M.S. (2020b) Composizione
i aliana in diac onia: Le pa ole compos e
dell’i aliano nel quad o della Mo ologia delle
Cos uzioni. Be lin / Bos on: De G uy e .
doi:10.1515/9783110652161.
Micheli, M.S. (2022) CODIT. Anew esou ce
o he s udy o I alian om adiach onic
pe spec i e: Design and applica ions in he
mo phological ield. Co pus 23. doi:10.4000/
co pus.7306.
Noailly, M. (1990) Le subs an i épi hè e
(Linguis ique nou elle). 1 e éd. Pa is: P esses
uni e si ai es de F ance.
Radimský, J. (2015) Noun+Noun compounds
in I alian: aco pus-based s udy (Epis eme.
Theo ia). Vydání 1. České Budějo ice:
Jihočeská uni e zi a Českých Budějo icích.
Radimský J. (2018) In lec ion o binominal
ATAP compounds in F ench and I alian:
apa adigma ic accoun . Lingue elinguaggio 2,
261–272. doi:10.1418/91868.
Radimský, J. (2022) Towa ds adiach onic
analysis o Romance mo phology h ough
Google n-g ams. Co pus 23. doi:10.4000/
co pus.7205.
Raine , F. (2012) Esca ceos sob e la con e sión
sus an i o → adje i o en español. In:
Fáb egas, A., E.Feliú, J.Ma ín and J.Pazó
(eds) Los lími es de la mo ología: Es udios
o ecidos aSoledad Va ela O ega, 369–382.
Mad id: UAM Ediciones. doi:10.15366/l.
mo ologia2012.022.
Raine , F. (2021) Compounding: F om La in
o Romance. Ox o d Resea ch Encyclopedia
o Linguis ics. Ox o d Uni e si y P ess.
doi:10.1093/ac e o e/9780199384655.013.691.
[las accessed 01/08/2023].
Sal i, G. and L.Vanelli (2011) Nuo a g amma ica
i aliana (Manuali Linguis ica). Bologna: Il
Mulino.
Scalise, S. and A.Bise o (2009) The
Classi ica ion o Compounds. In: Liebe ,
R. and P.Š ekaue (eds) The Ox o d
Handbook o Compounding, 34–53. Ox o d:
Ox o d Uni e si y P ess. doi:10.1093/
ox o dhb/9780199695720.013.0003.
Tho n on, A.M. (2004) Con e sione. In:
G ossmann, M. and F.Raine (eds) La
o mazione delle pa ole in i aliano, 499–533.
Tübingen: Max Niemeye Ve lag.
T augo , E.C. and G.T ousdale (2013)
Cons uc ionaliza ion and cons uc ional changes
(Ox o d S udies in Diach onic and His o ical
Linguis ics). Fi s edi ion. Ox o d: Ox o d
Uni e si y P ess.
Zanche a, E. and M.Ba oni (2005) Mo ph-
i ! A ee co pus-based mo phological
esou ce o he I alian language. P oceedings
om he Co pus Linguis ics Con e ence Se ies
1(1). A ailable a h ps://www.bi mingham.
ac.uk/ esea ch/ac i i y/co pus/publica ions/
con e ence-a chi es/2005-con -e-jou nal.
aspx [las accessed 01/08/2023].
Zinga elli (2011) Zinga elli Nicola, Lo Zinga elli
2011, Vocabola io della lingua i aliana,
CD-ROM. Bologna: Zanichelli.
Jan Radimský
Depa men o Romance S udies
Facul y o A s, Uni e si y o Sou h Bohemia in České Budějo ice
B anišo ská 31a, 37005 České Budějo ice
ORCID ID: 0000-0002-9016-9293
[email p o ec ed].cz
OPEN
ACCESS
JAN RADIMSKý 153
APPENDIX
Lis o N2s (modi ie s) in he sample, o de ed acco ding o dec easing ype equency
N2 English
gloss (N2)
Example
N1
English gloss
(N1N2)
Type e-
quency
Fi s ap-
pea ed
in
Las ap-
pea ed
in
chia e key
impo an pa ola wo d+key
keywo d 559 1806 2020
pilo a
pilo
leading, i-
al p oge o p ojec +pilo
pilo p ojec 153 1806 2020
an asma
ghos
imagina y,
alse ci à own+ghos
ghos own 140 1804 2020
ca dine
hinge
undamen-
al, co ne -
s one
pun o poin +hinge
linchpin 136 1830 2020
eco d eco d li ello le el+ eco d
eco d le el 83 1869 2020
leade leade azienda company+leade
leading company 64 1867 2020
gioca olo oy
alse pis ola gun+ oy
oy gun 62 1869 2020
guida guide linee lines+guide
guidelines 57 1802 2020
be saglio a ge o gano o gan+ a ge
a ge o gan 52 1819 2020
simbolo symbol luogo place+symbol
symbolic place 49 1805 2020
sa elli e sa elli e paese coun y+sa elli e
sa elli e coun y 47 1825 2020
quad o ame
gene al acco do ag eemen + ame
amewo k ag eemen 42 1804 2020
pi a a
pi a e
alse, aud-
ulen , ag-
g essi e
copia copy+pi a e
pi a e copy 38 1838 2020
ampone bu e
p o isional zona zone+bu e
bu e zone 38 1876 2020
lampo ligh ning
quick, sho gue a wa +ligh ning
bli zk ieg 37 1813 2020
p incipe
p ince
main, lead-
ing s umen o ins umen +p ince
main ins umen 36 1802 2020
OPEN
ACCESS
154 LINGUISTICA PRAGENSIA 2/2023
N2 English
gloss (N2)
Example
N1
English gloss
(N1N2)
Type e-
quency
Fi s ap-
pea ed
in
Las ap-
pea ed
in
an asia an asy
lashy essu o ab ic+ an asy
pa e ned ab ic 27 1836 2020
cul cul ilm mo ie+cul
cul mo ie 27 1881 2020
spazza u a ash ele isione ele ision+ ash
ash ele ision 25 1900 2020
pa ne pa ne paesi coun ies+pa ne
pa ne coun ies 23 1882 2020
spe acolo show en e -
ainmen poli ica poli ics+show
showbiz poli ics 21 1809 2020
omb a shadow go e no go e nmen +shadow
shadow go e nmen 19 1805 2020
pon e b idge
empo a y p es i o loan+b idge
b idge loan 19 1824 2020
ci e a
owl
coque e, de-
coy au o ca +owl
decoy ca 17 1804 2020
iume
i e
excessi ely
long disco so speech+ i e
excessi ely long speech 16 1861 2020
s imolo igge , in-
cen i e ase sen ence+ igge
igge sen ence 15 1816 2020
cuscine o cushion
bu e zona zone+bu e
bu e zone 13 1902 2020
ma igna
s epmo h-
e unkind,
wicked na u a na u e+s epmo he
wicked na u e 12 1802 2020
e iccio e ish
idol a o e ac o + e ish
e ish ac o 12 1828 2020
do mi o io do mi o y qua ie e dis ic + do mi o y
do mi o y dis ic 12 1895 2020
an occio puppe egime egime+puppe
puppe egime 11 1811 2020
boia
execu ion-
e
ex eme, in-
human
eddo cold+execu ione
eezing cold 10 1868 2020
agno spide uomo man+spide
spide man 8 1802 2020
cane dog
damn eddo cold+dog
eezing cold 6 1801 2020
OPEN
ACCESS
JAN RADIMSKý 155
N2 English
gloss (N2)
Example
N1
English gloss
(N1N2)
Type e-
quency
Fi s ap-
pea ed
in
Las ap-
pea ed
in
o za
s eng h
basic, un-
damen al idea idea+s eng h
basic idea 5 1804 2020
bomba bomb
sensa ional no izia news+bomb
bombshell 5 1885 2020
lumaca snail
e y slow eno ain+snail
e y slow ain 4 1809 2020
o ocopia pho ocopy
iden ical gio nali
newspape s+
pho ocopy
iden ical newspape s
( ea u ing he same
con en )
4 1869 2020
ca enaccio bol
blocking legge
law+bol
dec ee-law (adec ee
en e ing in o ce imme-
dia ely)
4 1884 2020
ghe o ghe o qua ie e dis ic +ghe o
ghe o dis ic 4 1938 2020
abocche o ap domanda ques ion+ ap
ick ques ion 3 1929 2020
ago needle
e y hin pesce ish+needle
needle ish 2 1802 2020
lage
Nazi con-
cen a ion
camp
manico-
mio
madhouse+
concen a ion camp
asylum whe e ill- ea -
men is in lic ed
2 1966 2020
a en u a ad en u e iaggio ip+ad en u e
ad en u e ip 2 1976 2020
ancobollo
pos age
s amp
e y small o ma o
o ma +pos age
s amp
pos age s amp o ma
1 1872 2020
ca ia guinea pig,
lab animal ci à own + guinea pig
expe imen al own 1 1956 2012
panino sandwich gio nali
newspape s+
sandwich
newspape ha is sold
oge he wi h ano h-
e newspape o mag-
azine
1 1976 2010
OPEN
ACCESS