Leita
Leita í textum Íslenska málfræðifélagsins    Leita í öllu efni á Kvisti

Um Moldhaugnaháls út í Fjósa og Fjörður

Af áhrifsbreytingum í nokkrum fleirtöluörnefnum

Útdráttur

A number of Icelandic plural place names show morphological developments that cannot be seen among ordinary nouns. In the paper, five categories of such developments are discussed, starting with an overview of the data in section 2:

(1) Masculine plural place names become feminine (2.1): three types can be distinguished, as illustrated in table 1; the largest one comprises masculines with nom. plur. -ir, acc. -i becoming feminines with nom. plur. -ir, acc. -ir.

(2) Feminine plural place names become masculine (2.2): the data falls into four categories, as shown in table 2; the most prominent one with feminines in nom.-acc. plur. -ar becoming masculines in nom. plur. -ar, acc. -a.

(3) Original neuter plural place names (2.3): four categories are distinguished in table 3, the largest one containing original neuters with zero ending (-ø) in nom.-acc. plur. becoming masculines with nom. plur. -ar, acc. -a.

(4) Plural place names based on u-stem nouns (2.4): a small body of data where original masculines in nom. plur. -ir, acc. -u (or younger acc. -i) become feminines in nom.-acc. plur. -ur.

(5) The development of the gen. plur. ending -na (2.5): in place names, the gen. plur. ending -na shows a tendency to spread beyond its original domain in weak feminine and neuter nouns. On the one hand, the gen. plur. ending -na spreads to words of other inflectional classes, especially in compounds, as shown in (17–19), or, less frequently, in non-compounds, exemplified in (20). On the other hand, the -n- of this gen. plur. ending can be shown to have penetrated other case forms of the plural paradigm, as shown in (21).

These changes fall into two main categories, as described in 2.6: on the one hand those that involve gender switch (changes in categories 1 through 4) and, on the other hand, the spread of the gen. plur. ending -na (category 5). The earliest of these changes (category 1) is in evidence already at the end of the fourteenth century.

At least two characteristics of place names separate them from ordinary nouns, as discussed in section 3: First, Icelandic place names are primarily used in the dative case, whereas the nominative is the most prominent case form of ordinary nouns. This is shown by statistics for Modern Icelandic in table 4, and a study by Jóhanna Barðdal (2001) indicates that the situation was not significantly different in Old Icelandic. This is discussed in section 3.1. Second, Icelandic place names frequently lack clear marking for gender: place names are rarely used with adjectives or a suffixed article that could reveal the gender (although there are exceptions). Moreover, the attributive adjectives most commonly used with place names (Ytri-, Innri-, etc.) could be either masculine, feminine or neuter. This is the topic of section 3.2. The effect of these two characteristics of place names is particularly profound in plural place names, since the principal case form, the dative plural, has a single morpheme, -um, in all three genders, across all inflectional classes.

Place names show analogical tendencies that are different from ordinary nouns, as discussed in section 4. Many of the general tendencies detected in analogical change by Kurylowicz (1945–49) and Manczak (1958, 1963) find a striking correspondance in data from language acquisition, as shown by Hooper (1980:166): forms that are acquired late (derived forms) frequently are affected by forms that are acquired early (basic forms). In ordinary nominal paradigms, the nominative case is basic or unmarked, but as Ma´nczak (1958:396–401) and Tiersma (1982) have pointed out, in place names the local case, dative in Icelandic place names, is the basic or unmarked case; this Tiersma refers to as local markedness.

The morphological developments described in section 2 are analyzed in 5. It is argued that these changes all are due to local markedness (the dative is the unmarked case form in place names) and the fact that the unmarked case form is ambiguous with respect to gender (dat. plur. has a single morpheme -um in all three genders, across all inflectional classes). The changes that involve gender switch (5.2) seem to suggest that the reanalysis triggered by the gender-indeterminate dative forms is to some extent governed by lexical strength, as defined by Bybee (1985): masculines in nom. plur. ir, acc. -i tend to join the more numerous and more productive feminines in nom.-acc. plur. -ir (5.2.1), feminines in nom.-acc. plur. -ar tend to enter the larger category of masculines in nom. plur. -ar, acc. -a (5.2.2), the neuters with zero ending

(-ø) in nom.-acc. plur. tend to join the large masculine class in nom. plur. -ar, acc. -a (5.2.3), and the small class of u-stem place names in nom. plur. -ir, acc. -u (or younger -i) tend to join the more numerous feminines in nom.-acc. plur. -ur (5.2.4). The uniform morphology of the dat. plur. (-um) across all inflectional classes allows the gen. plur. ending -na to spread by analogical extension (5.3).