Íslenskt mál og almenn málfræði - 01.01.2021, Page 156

pounds:16 they produce determinative compounds whose head further clarifies
the type of geographic entity (e.g. Jórsalaborg). These compounds have the same
denotation as the modifier alone. Finally, 4) independent native neoformations
(e.g. Mikligarðr), which do not depend on foreign models.
Latin integral borrowings (121 in total) are worth looking into in order to
investigate their status in the language. On the one hand they are not integrated,
if not (only occasionally) orthographically (e.g. apokalipsis), and often show Latin
inflection. On the other hand they may be acclimatized, which is the process
whereby a lexeme becomes an active part of the lexicon. Acclimatization and
integration do not always go hand in hand. Thus, one does not exclude the other
(cf. above, § 2.3). In the corpus, Latin integral loans are almost equally divided
into nonce borrowings, thus words which do not have an active status in the lex-
icon (e.g. antipes, ebdomada, periphrasis), and independent lexemes, active mem-
bers of the lexicon (e.g. Sagittarius, rhetorica, studium). This latter category is
slightly more common in the excerpted data (66 lexemes). Finally, four “loans”
(dominus, filius, magister and spiritus sanctus) pertain to scribal abbreviations, one
of the two outlier categories according to the taxonomy outlined above (§ 2.3).

6. Word pairs from 1550 onwards
According to Gusmani’s (1981:145–148 and 157–167) theoretical model, the coex-
istence of loanwords and native words produces an unbalance in the lexicon. In
order for the balance to be re-established, two chief processes are found: neutral-
ization and polarization. Neutralization involves the elimination of either loan-
word or native word from the lexicon, preceded by a period of competition during
which either of them comes to be prominently used. Polarization is the process
whereby either of the two words becomes differentiated from the other to some
extent, for example semantically, diastratically or diaphasically.
Matteo Tarsi156
The coining of OIcel. Jórsalir is best explained by taking into account two principles:
phonological resemblance and semantic motivation. The former seems however to
have played a smaller role than the latter (cf. Mod.Icel. Jerúsalem vs. OIcel. Jórsalir).
OIcel. Jórsalir is semantically motivated in Icelandic by OIcel. jór- = jǫfurr ‘king,
chieftain’ (ÍOb, s.v. jór-) and -salir, plural of OIcel. salr ‘residence, home’ (< PGmc
*sali-), which occurs in native place-names (e.g. Uppsalir ‘Uppsala’). Cf. also Lgb.
*sala in many toponyms throughout the Italian peninsula and It. sala ‘living room’ (<
Lgb. *sala ‘household, court’). The formation of OIcel. Jórsalir thus accomodates a
twofold need: on the one hand the Latin place-name is reanalyzed according to a
word formation pattern common in Old Norse place-names, while on the other it
acquires transparency, regardless of its original meaning.
16 The term “epexegetical compound” was adopted from Alexandra Petrulevich’s work
(see e.g. Petrulevich 2016 and 2017).